Silver Creek Dam Potential Failure Modes Analysis (PFMA)

Final Report
Silver Creek Dam Potential Failure Modes

Analysis (PFMA)

11 APRIL 2011

EXECUTIVE SUMMARY
E.1 Introduction
Silver Creek Dam is located on Silver Creek, approximately 2 miles southeast of Silverton, Oregon. The
drainage area for the dam is approximately 46 square miles and extends southeast of the project and
includes Silver Falls State Park. The dam was designed by CH2Mhill and constructed between May 1973
and November 1974 for the City of Silverton, Oregon. Due to the history of seepage and proximity of
the dam to residents of Silverton downstream of the dam, the City has decided to construct an
automated data collection and early warning system in order to try to improve its ability to monitor the
condition of the dam and to provide warning to residents if unsafe conditions develop at the dam.
The City of Silverton and the Portland District US Army Corps of Engineers (USACE) have agreed to
partner on the design and installation of the data collection and warning system as part of the
Continuing Authorities Program (CAP). As part of this project, USACE recommended that a potential
failure modes analysis (PFMA) be conducted for the dam. The purpose of the PFMA was to identify
potential modes of failure for the dam in order to evaluate what measures may be taken to reduce
either a) the likelihood that failure of the dam will occur and/or b) the consequences if the dam should
fail. The results of the PFMA will be used to focus the scope of work for the design of the monitoring
and early warning system.
E.2 Results of PFMA
The PFMA team, consisting of engineers from USACE and City Operations Personnel identified 36
potential failure modes. Of these, 16 were preliminarily identified as being credible and significant.
After evaluating these 16 failure modes further, some of the failure modes were found to be similar and
were combined and some were considered credible but not significant. After further evaluation, 4
failure modes were considered credible and significant. The credible and significant failure modes were
identified as: Spillway Slab Jacking, Downstream Sill and Stilling Basin Scour, Upstream Sill Scour and
Overtopping. Seepage was considered, however it was judged to be less of a concern than the four
significant failure modes identified.
E.3 Major Findings and Understandings
During the PFMA session the following Major Findings and Understandings were identified:
1) The Dam appears to be well maintained and project operations staff is maintaining the project to
the best of their abilities.
2) A variety of documents were available for review for the PFMA. A list of the documents is included
as Appendix A. In general, there was a lack of construction photos associated with original
construction (one aerial photo was available from original construction).
3) General access to the right side of the project is by car. Access to the right abutment of the project
site (including the right spillway training wall and fish ladder) can be gained on foot. Access to the
embankment, low level outlet and left abutment of the dam requires a boat.
ES‐1

4) Under normal operations, pool rises to spillway crest (El. 424 feet) for about nine months of every
year. Pool of record is about El. 430 feet (6 feet over the crest) .
5) Downstream flooding occurs every year at Silver Gardens Retirement Home, requiring sand bagging
every year. This will continue unless structural measures are taken to prevent flooding.
6) There is no stream gauging upstream of the dam or forecast data available at the project. As a
result, the City is unable to predict flooding in advance from rainfall.
7) There are no weirs to measure seepage present at this project. The city currently manually
measures seepage with a bucket and stopwatch periodically. This method results in inconsistent
data and no plotting of seepage data has occurred to date.
8) There is no seepage collection/measuring present on the left abutment. Because of the historical
seepage concerns, seepage was extensively discussed. For this PFMA, general seepage was broken
down into seepage leading to a specific failure mechanism (i.e. increased seepage causes an
increase in uplift pressure beneath the spillway slabs, leading to failure of spillway slabs, which
allows erosion of the underlying foundation materials, leading to breach of the dam).
9) Outlet capacity and reliability are keys to safe project operations. The project cannot dewater
quickly if needed for an emergency (except in summer). In addition, there are reliability concerns
based on operation history.
10) Inspectors/City Staff cannot view or inspect the left side of the Dam (embankment, spillway and
abutment) without a boat. This results in infrequent and inadequate inspection.
11) City staff cannot intervene to stop dam failure without the use of a helicopter or boat/barge.
Placing material at the dam, such as rock or other materials to prevent or stop dam failure is
difficult.
12) The current Emergency Action Plan (EAP) does not allow for full contact with the population. The
call list for the current emergency public notifications system (Everbridge Call System) is not
complete. According to City Operations Staff, the suppliers of the old notification system (Code
Red) will not release the full list to the City. The Everbridge system will automatically call all
landlines in Silverton. However, City residents won’t be contacted on cellular phones unless they
have signed up for the Everbridge call system (even if they signed up originally for the old Red Code
system).
13) The Project Operations plan needs trigger thresholds for inspection (when to go to 24‐hour
surveillance, etc.) and warning. Critical elevations need to be established for this project in order
to ensure adequate inspection and early warning.
14) Known foundation features will continue to seep (basalt interflow zone @ 398, landslide talus on
right abutment). However, these areas are not prone to piping and dam failure.
15) Visible signs of distress are present at the dam (the spillway training walls are tilted, stilling basin
scour is apparent at the site, new and/or increased seepage has been identified in recent years).
16) The City reports that seepage has increased in recent years (left and right side of outlet works, new
pipe installed in Feb 2010), but there is no way for the City to quantify or measure to verify if
seepage is increasing due to a lack of a weir or other measurement system that could be used to
quantify this seepage.
17) The City inspects the dam twice per year and the State inspects the dam approximately once per
year. Inspection requires boat access so frequent inspection is difficult.
ES‐2

18) Stilling basin scour is evident and has been previously reported in post construction inspections.
There is no record of scour inspection or monitoring by the City. Original recommended operations
and maintenance document indicates that the stilling basin is prone to scour and should be
frequently inspected. The inspections will require diverting water from the low level outlet and fish
ladder and pumping from the stilling basin during periods of no flow through the spillway; or use of
an ROV.
19) Power at the project is City supplied and is only currently available on the right side of the Dam. No
backup power, such as an emergency generator, is available at the site. Some method of
alternative power, such as a solar panel, may be needed for any proposed monitoring system.
20) Some of the piezometers are currently non‐functional and some of the piezometer standpipes are
small in diameter. Small diameter piezometers may prove difficult to automate. If new
piezometers are required, access to the embankment will have to be gained by boat or helicopter.
21) Operations personnel have no means of verifying the low level outlet gate position (42‐in. or 18‐in
outlet) – operations cannot see the outlet from the control stand located on the left bank.
22) Outlet operation during high flows (open or closed) is not defined in the operations manual. The
outlet may not be structurally sound at high pools when the gate is open due to the potential for
cavitation. In addition, the air intake for the outlet may be inundated in high flow events which will
result in surge flow.
23) Low spots are present at the right abutment. This means overtopping may occur at lower
elevations.
24) No current hydraulic modeling/routing has been performed in order to define thresholds (100 yr
flood, PFM, etc.), pools and tailwaters. If modeling data was available, the ability to set meaningful
thresholds for an early warning system would be improved.
25) While the top of the embankment is designed at elevation 440, the clay core within the
embankment only extends up to Elev. 435; the filter extends up to Elev. 430. Increased seepage
above the clay core should be anticipated with pool levels greater than El. 435. In addition, the lack
of a clay core extending to the top of the dam may result in overtopping occurring at elevations
below elevation 440 (i.e. overtopping may begin before the pool reaches the top of the dam due to
wave/wind erosion of the embankment). The potential for overtopping to begin at a lower
elevation should be considered when defining trigger points for an early warning system.
26) The City was not able to produce current survey data for true height of embankment crest, the tops
of walls or abutments. Settlement of the embankment or other low spots may result in
overtopping at lower elevations than anticipated. Knowing the location of low spots is important
for determining locations of stocking emergency materials and to focus flood fighting efforts.
27) Based on the results of analysis performed to date by Cornforth Consultants, the embankment and
abutments are expected to be adequately stable under seismic conditions and liquefaction
susceptibility is low.
28) No seismic analysis of the concrete spillway or spillway wall has been performed. It is anticipated
that loss of the wall may lead to failure of the embankment. (Walls designed for 0.05g, current MCE
from USGS = 0. 4g).
29) There appears to be a gap in the core trench on the right abutment below the training wall. The
presence of a gap at this location may result in seepage at pools above Elev. 430.
ES‐3

30) Intervention materials (sand bags, rip rap) are not currently stockpiled on the site.
31) There are residence locations within very short travel times downstream of the dam. Warning time
would not be possible for some failure modes.
32) Loss of pool in summer months will eliminate the downstream water supply and force the City to
go to water conservation.
33) A log boom was installed at the end of construction at the upper end of the pool but is no longer
present. Build up of reservoir rim debris could clog the spillway during high flow events.
34) Large vegetation debris is expected to continue to collect downstream of the spillway below the
stilling basin, reducing channel capacity. If this material continues to collect over many years, it
may create a downstream hazard during a subsequent large flood event as occurred in recent
flooding events on the Sandy River resulting in downstream damage to bridges and property.
35) Project staff was unaware of the presence of 4 foot filter drain that runs from the filter zone to the
left of the outlet pipe and ends downstream immediately to the left of the outlet structure (this
may or may not exist since as‐built drawings do not show it but it is discussed in CH2MHill
correspondence). If it exists, this drain may contribute to seepage exiting near the low level outlet
structure at the toe of the dam.
36) There are no records of conduit inspection/ROV inside the outlet pipe.
37) Zone 1 materials on the upstream shell of the embankment consist of well graded silty, sandy
gravel and should act as a flow limiter/piping restriction (fill developed voids) should it initiate.
38) The low level conduit passes through the dam at the base of the embankment. It is founded on
bedrock where it passes beneath the impervious core. The conduit is square and has seepage
collars on 10 foot centers. Impervious back fill around the conduit was placed in 4 inch lifts and
compacted with hand tampers.

E.4 Conclusions
1) Failure modes related to seepage through the right abutment landslide material are considered
credible but not as significant as the 4 credible and significant failure modes identified in this report.
It was judged that even under high pool conditions, such as Elev. 440 ( PMF conditions), the relative
increase in hydraulic gradient compared to what the project has experienced is not enough to
initiate piping in the generally pervious, angular, basalt fragments that comprise the landslide
deposit. However, it is important to continue monitoring seepage to detect any increases that may
indicate a potential developing issue of concern.
2) With respect to the significant and credible failure modes, an automated early warning system
would be most effective at reducing risk associated with the overtopping failure mode.
3) For overtopping failure modes, it is anticipated that, if properly designed and maintained, an
automated early warning system may be able to provide early warning to decision makers in time to
reduce the risk associated with overtopping failure modes.
4) Initiation and progression of the failure modes associated with erosion of the spillway foundation
would be difficult to monitor with an automated monitoring system and therefore difficult to
ES‐4

provide advanced warning of failure by an early warning system. Monitoring horizontal drain data
and data from a new piezometer installed to the left of the left spillway wall and downstream of the
embankment filter zone could be used to monitor development of the failure mode associated with
upstream sill scour and increased seepage through the right abutment. However, visual monitoring
for spillway slab offsets and stilling basin and upstream sill erosion are more direct ways to identify
whether these failure modes are initiating.
5) For failure modes associated with spillway foundation erosion, it is anticipated that the time from
initiation to uncontrolled release of the pool could be months if the failure occurred under normal
pool conditions but days (possibly hours) under high flood (such as a 100 year event). During high
flows, unless initiation is observed directly by City staff, minimal warning will be provided to the
downstream population for spillway failure initiated failure modes regardless of whether a
warning system is installed.
6) If properly designed and maintained, the automated data collection function of the proposed early
warning system will provide a benefit of increased consistency and quality of data (including
piezometer and drain data) collected by the system. All data should be stored for future reference.
This data helps to assess the condition of the dam and improves understanding for how the dam
responds under various pool and seasonal conditions.
7) The automated data collection function could also be used to provide confirmation of flows through
the low level outlet and frequent updates on the elevation of Silver Creek Dam downstream of the
dam to improve coordination of flood response activities associated with hydrologic events that are
not related to dam failure.
E.5 Recommendations Regarding the Early Warning System
1) The automated early warning system should include a reservoir level measurement and rain gage.
Multiple level thresholds for reservoir level and rain data should be defined as part of the design of
the early warning system. At a minimum, the thresholds should consider the elevations of the top
of the core (El. 435) and the low spot to the right of the right abutment wall (approx. El. 438). In
addition, due to the gap in the core trench on the right abutment below the training wall, the City
should anticipate the possibility of increased seepage with pools above Elev. 430 and monitor this
location. The system should provide notification of the exceedance of the thresholds to appropriate
decision‐making personnel at the City. In addition, it should collect and store data for future
reference and analysis.
2) The City should design and install weir collection boxes to measure the flow rate of known seeps
and drain flows, especially flows associated with the toe drains. Changes in these flow rates may
indicate changing conditions at the dam.
3) In the event that drain flows show sustained increases in flow, consider installation of a new
piezometer to the left of the left spillway wall for the purposes of monitoring piezometric pressure
within the landslide deposit that provides the foundation for the spillway. The water level in this
ES‐5

piezometer could be monitored with an automated data collection system with thresholds assigned
to provide notification.
4) The reliability of the existing piezometers should be evaluated as part of the design of the warning
system.
5) For the purposes of historical data collection, the automated warning system should be designed to
collect data from the piezometers and weirs to improve the consistency of data collection and
review.
6) The automated system should support storage of historical data for the purposes of reviewing
historical trends and providing reports when needed. Time history plots of data should be reviewed
at least monthly by the City and more frequently if visual observations indicate potential developing
issues of concern.
7) For the purposes of improving flood response by the City, the automated warning system should be
designed to collect data for water level at downstream locations as determined by the City.
Multiple level thresholds can be developed to provide an assessment of level of urgency.
8) The City’s Emergency Action Plan should be updated to incorporate the early warning system and to
address notification of all citizens impacted by flooding.

E.6 Additional Recommendations to Improve Dam Safety and Reliability
1) Perform a survey of the dam along the crest of the dam from left abutment to right abutment in
order to determine the presence of low points on the dam. If there are low spots, they should be
raised to the design elevation.
2) Perform survey of spillway and abutment walls for vertical and horizontal deflection in order to
establish a baseline for comparison with future surveys. Additional surveys should be completed at
least every 5 years. If surveys indicate movement, the results of the surveys should be reviewed by
a structural engineer for evaluation.
3) Inspect and survey the stilling basin for scour during period when spilling is not occurring (summer
low pool inspection) in order to establish a baseline for comparison with future surveys. Inspect the
stilling basin for increased scour after large flow events. Perform the inspections and surveys at
least every 5 years.
4) Inspect the spillway for vertical offsets between slabs during period when spilling is not occurring
(summer low pool inspection). The most critical offset is the downstream slab uplifted relative to
the upstream slab; this creates localized high pressures that may penetrate down along the slab
joint and create high uplift pressure. If downstream vertical offsets are present, grind the
downstream slab down to remove offsets. Recommend visual inspections are completed at least
once per year.
5) Visually inspect upstream impervious blanket for holes for its full length upstream (not just the
upstream sill) when possible. Visually inspect sillfor erosion at least once per year or after large spill
events.
6) Inspect the integrity of the low level outlet pipe using a Remotely Operated Vehicle (ROV).
ES‐6

7) Evaluate and implement options for increasing the reliability of the low level outlet hydraulic system
and for confirmation of gate opening size. To this end, the warning system should include collection
of data on the flow through the low level outlet to improve the ability to confirm gate opening size.
8) The City’s new public notification system (Everbridge) will call all City of Silverton telephone
landlines but does not include the entire contact list from the provider of the old public notification
system (Code Red). Recommend evaluating ways to update the Everbridge system contact list for all
cellular phone users to the extent possible.
9) Perform hydraulic modeling to determine the frequency and magnitude and routing of storm events
(determine relationships between flood frequency, peak flow and downstream water elevations).
The results of the modeling would help the City with decisions regarding balancing downstream
flooding impacts with dam safety risks and to improve the City’s ability to warn the population.
Once the results of the modeling are complete, it may be appropriate to modify the operations plan
for the project.
10) The City should consider making structural improvements to prevent flooding of residents at low
elevation locations (such as Silver Gardens Retirement Home).
11) The increased woody debris building up downstream of the stilling basin creates a potential for
downstream damage during future storm events. During large events, this material may become
dislodged and cause damage to bridges and property. The City should consider removing this
material every 5 years.
ES‐7

Table of Contents
1.0 INTRODUCTION ...................................................................................................................................... 1
2.0 BACKGROUND ........................................................................................................................................ 1
3.0 PFMA PROCESS ...................................................................................................................................... 2
4.0 MAJOR FINDINGS AND UNDERSTANDINGS ............................................................................................ 3
5.0 CREDIBLE AND SIGNIFICANT FAILURE MODES ........................................................................................ 6
5.1 Spillway Slab Jacking (PFM C in Appendix B) ...................................................................................... 6
5.1.1 Description ................................................................................................................................... 6
5.1.2 Factors that Increase the Likelihood of this PFM ......................................................................... 7
5.1.3 Risk Reduction Measures ............................................................................................................. 7
5.1.4 Warning Time ............................................................................................................................... 7
5.2 Downstream Sill and Stilling Basin Scour (PFM D and PFM E in Appendix B) .................................... 7
5.2.1 Description ................................................................................................................................... 7
5.2.4 Warning Time ............................................................................................................................... 8
5.3 Scour of upstream spillway sill (PFM F in Appendix B). ...................................................................... 8
5.3.1 Description ................................................................................................................................... 8
5.3.4 Warning Time ............................................................................................................................... 9
5.4 Overtopping (PFM J, K and N in Appendix B). .................................................................................... 9
5.4.1 Description ................................................................................................................................... 9
5.4.4 Warning Time ............................................................................................................................. 10
5.5 Right Abutment Seepage ................................................................................................................. 10
6.0 APPLICABILITY OF AN EARLY WARNING SYSTEM TO SIGNIFICANT FAILURE MODES ........................... 10
7.0 CONCLUSIONS AND RECOMMENDATIONS ........................................................................................... 12
7.1 Conclusions ....................................................................................................................................... 12
7.2 Recommendations ............................................................................................................................ 13
i

7.2.1 Automated Data Collection and Early Warning System Recommendations ............................ 13
7.2.2 Additional Recommendations to Improve Dam Safety and Reliability .................................... 14

Figures
Figure 1 Site Plan
Figure 2 Sketch of Spillway Slab Jacking Failure Mode
Figure 3 Sketch of Scour of Downstream Spillway Sill and Stilling Basin Failure Mode
Figure 4 Sketch of Scour of Upstream Spillway Sill Failure Mode
Appendices
Appendix A ‐ References
Appendix B – PFMA Notes
Appendix C – Site Visit Notes and Photographs
Appendix D – Project Drawings
ii

1.0 INTRODUCTION
Silver Creek Dam is located on Silver Creek, approximately 2 miles southeast of Silverton, Oregon. Silver
Creek is tributary to Pudding River approximately 6 miles below the project. The dam was designed by
CH2MHill and was constructed between May 1973 and November 1974. The dam was constructed to
store water for municipal water supply and recreation. Due to the history of seepage and proximity of
the dam to residents of Silverton downstream of the dam, the City of Silverton has decided to construct
an automated data collection and early warning system in order to try to improve its ability to monitor
the condition of the dam and to provide warning to residents if unsafe conditions develop at the dam.
The City of Silverton and the Portland District US Army Corps of Engineers (USACE) have agreed to
partner on the design and installation of the data collection and warning system as part of the
Continuing Authorities Program (CAP). As part of this project, USACE recommended that a potential
failure modes analysis (PFMA) be conducted for the dam. The purpose of the PFMA was to identify
potential modes of failure for the dam in order to evaluate what measures may be taken to reduce
either a) the likelihood that failure of the dam will occur and/or b) the consequences, if the dam should
fail. Failure is defined in USACE Engineering Regulation (ER) 1110‐2‐1156 – Safety of Dams – Policies and
Procedures and by the PFMA team as sudden, rapid and uncontrolled release of impounded water. A
failure mode is a detailed description of a possible chain of events from initiation through failure. Risk is
defined as the probability of a load (such as a pool level or earthquake) multiplied by the probability of
failure (given the load), multiplied by the consequences of the failure. As a result, the PFMA provides a
means for identifying risk at the project. The results of the PFMA provide a sense of what possible
modes of failure an automated early warning system might be effective at reducing risk but also for
those that the system may not be able to effectively reduce risk. The PFMA also provides valuable
information that can be used to determine what data should be collected from the automated data
collection and early warning system. Information learned from the PFMA can also inform the City of
critical aspects or areas on the project that require special attention.
2.0 BACKGROUND
The dam was constructed between May 1973 and November 1974. Upon first filling, seepage was
observed on the right side of the earth embankment. Figure 1 is a site plan for the dam. The reservoir
was lowered shortly thereafter and a series of horizontal drains were installed into foundation materials
below the right side of the earth embankment. In addition, a buttress filter layer berm was placed on
much of the right side of the embankment to intercept seepage not captured by the drains.
In June 1981, USACE and the Oregon Water Resources Department completed a Phase 1 Inspection
Report. It identified the Silver Creek Dam as a high hazard dam because of the potential loss of life risk
and the level of potential property damage immediately downstream of the dam. The inspection
evaluated abutment and foundation conditions, embankment stability, hydraulic and hydrologic
1

Right Abutment
Fish Ladder
Embankment
Spillway
Left Abutment
Figure 1. Site Plan

conditions, and structural/mechanical features. The inspection found the dam to be in satisfactory
condition for continued operation.
In March 1993, a magnitude 5.6 earthquake (Scotts Mills) occurred with an epicenter approximately 10
miles east of the City of Silverton. The magnitude of the earthquake and its location near the dam
caused increased focus on seismic safety in the region. In July 1999, the City contracted to Cornforth
Consultants to perform an assessment of the stability of the dam under steady‐state, rapid drawdown
and seismic conditions. The analysis indicated that the dam demonstrated adequate factors of safety
for each of the conditions considered.
At the request of the City of Silverton, a dam break analysis was performed by Philip Williams and
Associates in January 2000. The analysis was performed to provide information to be incorporated into
an Emergency Action Plan (EAP) for the project. The analysis concluded that full breach of Silver Creek
Dam would be catastrophic to the City and that elapsed time from dam failure to the passage of the
flood wave is less than one hour. It was recommended in this report that the City installs an early
warning system at the dam.
In 2002, Squier Associates performed a preliminary design report for an early warning system. The
design report included recommendations for an early warning system based on 3 failure mechanisms:
seepage failure under normal operating conditions, seismic deformations and instability and
overtopping failures resulting from flooding.
USACE regulations (ER 1110‐2‐1156) now require that a PFMA process be performed for all major dam
modifications or evaluation of interim risk reduction measures. In 2010, USACE decided that a PFMA
was necessary to evaluate the applicability of an early warning system to reduce risk at Silver Creek
Dam.
3.0 PFMA PROCESS
The first step in the process was to identify the team members that would participate in the PFMA. In
general, PFMA teams include a facilitator, co‐facilitator and technical experts. The makeup of the
technical experts on a PFMA team depends upon the nature of the project. Projects that have a high
degree of complexity may require team members from each of Structural, Geotechnical, Mechanical and
Electrical Engineering and Hydraulics and Hydrology disciplines in addition to members familiar with
operational aspects of the dam. For Silver Creek Dam, it was decided that the complexity of the project
with regard to electrical and mechanical equipment was such that team members representing the
other disciplines and the City’s operational team member have sufficient knowledge in this area to
address issues in these disciplines.
Once the team was identified, each of the team members reviewed background information on the
project. Project information was gathered by the City of Silverton and was transmitted to USACE prior
to the start of the PFMA, which was performed as a focused workshop, between 10 and 14 January,
2011. Each of the team members reviewed the background information prior to the PFMA workshop, so
that each member had familiarity with the project by the time of the workshop. On day 1, the team
2

continued to review background information as a group and summarized initial major findings and
understandings (see Section 4.0). On the second day, the team conducted a site visit to the project so
that the team could become more familiar with site conditions. The remainder of the week was spent
identifying, describing and evaluating potential failure modes for the project. The team identified the
major features of the project and based on the team’s experience and training for performing PFMAs
identified possible failure modes. With this list of potential failure modes, the team developed a
preliminary rating as to how significant each potential mode was. By convention, USACE assigns each
mode one of the following categories:
a) Credible: a physically plausible failure mode.
b) Significant (or Credible and Significant): these failure modes are a subset of credible failure
modes. The probability of failure and associated consequences approach closely or exceed a
tolerable risk limit guideline.
c) Non‐credible: not a physically plausible failure mode.
Appendix B includes the detailed notes that were taken during the workshop.
4.0 MAJOR FINDINGS AND UNDERSTANDINGS
This section includes a list of the major findings and understandings about the project that were
identified from review of the background information, the site visit and during the PFMA workshop:
1) The Dam appears to be well maintained and project operations staff is maintaining the project to
the best of their abilities.
2) A variety of documents were available for review for the PFMA. A list of the documents is included
as Appendix A. In general, there was a lack of construction photos associated with original
construction (one aerial photo was available from original construction).
3) General access to the right side of the project is by car. Access to the right abutment of the project
site (including the right spillway training wall and fish ladder) can be gained on foot. Access to the
embankment, low level outlet and left abutment of the dam requires a boat.
4) Under normal operations, pool rises to spillway crest (El. 424 feet) for about nine months of every
year. Pool of record is about El. 430 feet (6 feet over the crest) .
5) Downstream flooding occurs every year at Silver Gardens Retirement Home, requiring sand bagging
every year. This will continue unless structural measures are taken to prevent flooding.
6) There is no stream gauging upstream of the dam or forecast data available at the project. As a
7) There are no weirs to measure seepage present at this project. The city currently manually
8) There is no seepage collection/measuring present on the left abutment. Because of the historical
seepage concerns, seepage was extensively discussed. For this PFMA, general seepage was broken
3

down into seepage leading to a specific failure mechanism (i.e. increased seepage causes an
increase in uplift pressure beneath the spillway slabs, leading to failure of spillway slabs, which
allows erosion of the underlying foundation materials, leading to breach of the dam).
9) Outlet capacity and reliability are keys to safe project operations. The project cannot dewater
quickly if needed for an emergency (except in summer). In addition, there are reliability concerns
based on operation history.
10) Inspectors/City Staff cannot view or inspect the left side of the Dam (embankment, spillway and
11) City staff cannot intervene to stop dam failure without the use of a helicopter or boat/barge.
difficult.
12) The current Emergency Action Plan (EAP) does not allow for full contact with the population. The
call list for the current emergency public notifications system (Everbridge Call System) is not
complete. According to City Operations Staff, the suppliers of the old notification system (Code
Red) will not release the full list to the City. The Everbridge system will automatically call all
landlines in Silverton. However, City residents won’t be contacted on cellular phones unless they
have signed up for the Everbridge call system (even if they signed up originally for the old Red Code
system).
13) The Project Operations plan needs trigger thresholds for inspection (when to go to 24‐hour
surveillance, etc.) and warning. Critical elevations need to be established for this project in order
to ensure adequate inspection and early warning.
14) Known foundation features will continue to seep (basalt interflow zone @ 398, landslide talus on
right abutment). However, these areas are not prone to piping and dam failure.
15) Visible signs of distress are present at the dam (the spillway training walls are tilted, stilling basin
scour is apparent at the site, new and/or increased seepage has been identified in recent years).
16) The City reports that seepage has increased in recent years (left and right side of outlet works, new
pipe installed in Feb 2010), but there is no way for the City to quantify or measure to verify if
seepage is increasing due to a lack of a weir or other measurement system that could be used to
quantify this seepage.
17) The City inspects the dam twice per year and the State inspects the dam approximately once per
18) Stilling basin scour is evident and has been previously reported in post construction inspections.
There is no record of scour inspection or monitoring by the City. Original recommended operations
and maintenance document indicates that the stilling basin is prone to scour and should be
frequently inspected. The inspections will require diverting water from the low level outlet and fish
ladder and pumping from the stilling basin during periods of no flow through the spillway; or use of
an ROV.
19) Power at the project is City supplied and is only currently available on the right side of the Dam. No
backup power, such as an emergency generator, is available at the site. Some method of
4

20) Some of the piezometers are currently non‐functional and some of the piezometer standpipes are
small in diameter. Small diameter piezometers may prove difficult to automate. If new
piezometers are required, access to the embankment will have to be gained by boat or helicopter.
21) Operations personnel have no means of verifying the low level outlet gate position (42‐in. or 18‐in
outlet) – operations cannot see the outlet from the control stand located on the left bank.
22) Outlet operation during high flows (open or closed) is not defined in the operations manual. The
outlet may not be structurally sound at high pools when the gate is open due to the potential for
cavitation. In addition, the air intake for the outlet may be inundated in high flow events which will
result in surge flow.
23) Low spots are present at the right abutment. This means overtopping may occur at lower
elevations.
24) No current hydraulic modeling/routing has been performed in order to define thresholds (100 yr
flood, PFM, etc.), pools and tailwaters. If modeling data was available, the ability to set meaningful
thresholds for an early warning system would be improved.
25) While the top of the embankment is designed at elevation 440, the clay core within the
above the clay core should be anticipated with pool levels greater than El. 435. In addition, the lack
of a clay core extending to the top of the dam may result in overtopping occurring at elevations
below elevation 440 (i.e. overtopping may begin before the pool reaches the top of the dam due to
wave/wind erosion of the embankment). The potential for overtopping to begin at a lower
elevation should be considered when defining trigger points for an early warning system.
26) The City was not able to produce current survey data for true height of embankment crest, the tops
of walls or abutments. Settlement of the embankment or other low spots may result in
overtopping at lower elevations than anticipated. Knowing the location of low spots is important
for determining locations of stocking emergency materials and to focus flood fighting efforts.
27) Based on the results of analysis performed to date by Cornforth Consultants, the embankment and
abutments are expected to be adequately stable under seismic conditions and liquefaction
28) No seismic analysis of the concrete spillway or spillway wall has been performed. It is anticipated
that loss of the wall may lead to failure of the embankment. (Walls designed for 0.05g, current MCE
from USGS = 0. 4g).
29) There appears to be a gap in the core trench on the right abutment below the training wall. The
presence of a gap at this location may result in seepage at pools above Elev. 430.
30) Intervention materials (sand bags, rip rap) are not currently stockpiled on the site.
31) There are residence locations within very short travel times downstream of the dam. Warning time
would not be possible for some failure modes.
32) Loss of pool in summer months will eliminate the downstream water supply and force the City to
go to water conservation.
33) A log boom was installed at the end of construction at the upper end of the pool but is no longer
34) Large vegetation debris is expected to continue to collect downstream of the spillway below the
5

35) Project staff was unaware of the presence of 4 foot filter drain that runs from the filter zone to the
left of the outlet pipe and ends downstream immediately to the left of the outlet structure (this
may or may not exist since as‐built drawings do not show it but it is discussed in CH2MHill
correspondence). If it exists, this drain may contribute to seepage exiting near the low level outlet
structure at the toe of the dam.
36) There are no records of conduit inspection/ROV inside the outlet pipe.
37) Zone 1 materials on the upstream shell of the embankment consist of well graded silty, sandy
38) The low level conduit passes through the dam at the base of the embankment. It is founded on

5.0 CREDIBLE AND SIGNIFICANT FAILURE MODES
During the PFMA, the team identified 36 potential failure modes (PFMs). Of these, 16 were preliminarily
identified as possibly being credible and significant. After evaluating these 16 failure modes further,
some of the identified failure modes were found to be similar and combined and some were considered
credible but not significant. In the end, 4 failure modes were considered credible and significant. In this
section, each of these significant failure modes are described, along with possible actions or measures
that can be taken to reduce risk associated with the failure mode. The PFMA notes (Appendix B)
includes more detailed information regarding the evaluation of each failure mode.

5.1 Spillway Slab Jacking (PFM C in Appendix B)

5.1.1 Description
Spillway flow impacts on vertical offset in spillway slabs which result from uplift pressure or slab
settlement (either under long term static conditions or seismically induced). The most critical offset is
the downstream slab uplifted relative to the upstream slab, which creates localized high pressures that
may penetrate down along the slab joint and create high uplift pressure. Landslide foundation materials
are exposed initiating erosion. Exposed talus/landslide is highly erodible and scour will progress.
Continued loss of foundation material results in additional slab failures upstream and additional
foundation exposure and erosion. Erosion progresses to the upmost slab eroding down to basalt
resulting in uncontrolled release. Figure 2 is a sketch of this failure mode.
6

Figure 2. Spillway Slab Jacking Failure Mode
5.1.2 Factors that Increase the Likelihood of this PFM
 Joints are key‐shaped with pre‐molded joint‐filler between slabs with a plastic water‐stop
embedded approximately 3‐inches below top of spillway slab (no doweling or structural tie
between slabs).
 Construction final report documents difficulty with rock anchor installation due to poor rock
quality and saturated foundation (water in anchor holes).
 Visual observation during the site visit showed turbulence/flow disturbance on the left side of the
spillway at the toe.
 Spillway flows occur approximately 9 months out of the year.
 Pool was very calm immediately downstream of end of spillway then turbulent downstream of
calm area (possible deep pool).
 No data/inspection except visual flow disturbances documented during site visit.
 Drains are not inspected or cleaned.

5.1.3 Risk Reduction Measures
 At times of non‐spilling, inspect for offsets between spillway blocks.
 Grind slabs flush at joints.
 Anchor the slabs.
 Fill voids/scour holes with concrete (holes around and under).

5.1.4 Warning Time
 Warning opportunity time (time from ability to detect failure mechanism to initiation of the
mechanism) is 0 or negative. It can be seen after a slab has failed (maybe).
 Breach formation time (time from initiation of breach to loss of dam): for normal pools –
months, and for high flows greater than 100 yrs – days, maybe hours.

5.2 Downstream Sill and Stilling Basin Scour (PFM D and PFM E in Appendix
B)

5.2.1 Description
Spillway flow initiates erosion of the basalt and spillway stilling basin concrete. Erosion continues,
undermining the spillway concrete and foundation. Failure/undermining of the spillway foundation at
the toe results in loss of spillway slab support. Loss of slab support permits slab failure, exposing the
talus/landslide foundation. Exposed talus is highly erodible and scour will progress. Continued loss of
foundation material results in failure of additional slabs upstream and additional foundation exposure
and erosion. Erosion progresses to the upmost slab, resulting in all spillway slabs having been displaced
and eroding down to basalt resulting in uncontrolled release. Figure 3 is a sketch of this failure mode.
 Spillway was extended due to poor foundation conditions (talus)
7

Figure 3. Spillway Downstream Sill and
Stilling Basin Scour Failure Mode
 Oct. 1975 inspection mentions scour at the sill.
 Foundation bedrock was described as poor to fair quality.
 Original inspection manual indicated concern about erosion of the stilling basin and recommends
frequent inspections and monitoring for erosion. (Maximum historical pool of Elev. 430
corresponds to flow of approximately 5,000 cfs).
 Pool was very calm immediately downstream of end of spillway then turbulent downstream of
calm area (possible deep pool). This may be indicative of a possible deep pool created by scour.
 Visually monitor for scour and survey the downstream sill and stilling basin.
 Armor stilling basin.
 Anchor the spillway slabs.

5.2.4 Warning Time
 Warning opportunity time is 0 or negative. It can be seen after a slab has failed (maybe).
 Breach formation time for normal pools – months; and for high flows greater than 100 yrs –
days, maybe hours.
5.3 Scour of upstream spillway sill (PFM F in Appendix B).

5.3.1 Description
Scour present on the upstream sill of the spillway permits additional flow under the spillway. The
resulting flow overtaxes the spillway drain system creating full uplift on the spillway and slabs. Increased
uplift and cycles of loading displace spillway slabs (by offset, settlement, anchorage failure, etc.) resulting
in exposed highly erodible talus/landslide foundation. Erosion progresses through continued loss of slab
support and scour. Scour progresses upstream to full pool resulting in uncontrolled release down to
bedrock (approximately El. 375). Figure 4 is a sketch of this failure mode.
 Upstream impervious blanket has historically eroded and been repaired with pit run rock which
does not provide the same seepage cutoff that the original blanket would provide.
8

Figure 4. Upstream Spillway Sill
Scour Failure Mode.
 Each repair cycle indicates removal of impervious blanket, resulting in reduced scour but
increased flow beneath the spillway.
 Ensure spillway drains are effective.
 Ensure upstream clay blanket remains impervious.
 Anchor slabs if needed.
 Add piezometer adjacent to the spillway as an indicator of hydraulic pressure under the spillway
slabs. Calculate acceptable uplift pressure threshold at which slabs become dislodged and
monitor for these pressures.
5.3.4 Warning Time
 Warning opportunity time is 0 or negative. It can be seen after a slab has failed (maybe).
 Breach formation time for normal pools – months, and for high flows greater than 100 yrs –
days, maybe hours.
5.4 Overtopping (PFM J, K and N in Appendix B).

5.4.1 Description
Pool rises above the El. 435 (top of the clay core), wave action initiates erosion of sandy/gravel in the
embankment upstream shell and crest. Material sloughing upstream and downstream decreases the
crest width/seepage path leading to overtopping and continued overtopping leads to down‐cutting of
the embankment. Breach occurs, down‐cutting to the valley floor leading to uncontrolled release and
downstream consequences.
 Core only designed to El. 435.
 Historical debris blockage could raise the pool.
 Insufficient free board
 Project cannot pass the PMF (according to PMF estimated in 1981 USACE Report)
 Flashy pool (i.e. pool rises quickly with rain events due to a small drainage basin).
 Actual crest elevation varies and is unknown due to a lack of survey data. True crest of dam could
be lower resulting in more frequent overtopping at a lower than predicted level.
 Low spot occurs at right abutment at El. 438 would permit overtopping at lower elevations than
PMF or rest of project.
 Insufficient draw down capacity
 Landslide prone areas in the basin (surge)
 Raise low spots
 Armor the upstream face
9

 Armor the downstream face

5.4.4 Warning Time
 Warning opportunity time is hours but could be increased with an early/flood warning system..
 Breach formation time is approximately 2 hours.
5.5 Right Abutment Seepage
Seepage through the right abutment landslide deposit is considered an important issue but failure
modes associated with this condition were not considered as critical as the significant failure modes
identified above. The reservoir has generally filled up to the spillway crest (El 424 ft) each year. When
observed, seepage has been clear with no evidence of piping initiating. Also, infrequent observations at
higher pools above spillway (pool of record is 430 feet), does not provide any evidence of
turbidity/cloudy seepage to suggest that piping may have initiated or occurred. There is no reason to
believe that going from pool of record of El 430 to a 100‐year pool of El 431 (one additional foot higher
or 2 percent increase in depth of water down to the base of the embankment) would change the
hydraulic gradient sufficiently to initiate piping. A PMF event would raise pool 10 feet above pool of
record (20 percent higher). A 20 percent increase seepage gradient was thought not to be sufficient to
initiate piping of the pervious landslide debris. However, it is important to continue monitoring seepage
to detect any increases that may indicate a potential developing issue of concern.
6.0 APPLICABILITY OF AN EARLY WARNING SYSTEM TO SIGNIFICANT
FAILURE MODES
The term ‘early warning system’ can be interpreted to encompass a variety of systems that could
potentially be beneficial for reducing risk at this project. In a very simple form, an early warning system
could simply be a ‘flood warning system’ that utilizes rainfall and/or reservoir water level data to
provide notification of elevating pool conditions. In this simple form, the system would be pretty basic
and have only that function. In more advanced form, the system could monitor data from a variety of
instruments, compare that data to predetermined algorithms or thresholds based on several different
failure modes and provide notification when certain conditions are met. This type of system might be
referred to as an ‘early dam failure warning system.’ For the purposes of this report, it is assumed that
the City desires to install a system consistent with the latter description.
For failure modes that have parameters that are directly related to the initiation and further
development of those failure modes that can be monitored through automation, it is possible that a
system could be designed to provide early detection and notification of these modes of failure. The
effectiveness of this system would be rated based on the degree to which the initiation and
development of the most significant failure modes (i.e. those that have the highest likelihood of causing
significant downstream consequences) can be detected by the automated system in sufficient time to
both notify decision makers and to allow time for intervention and/or evacuation. Clearly, then, the
10

effectiveness of the system at a given project will depend upon the nature of the most significant failure
modes at that project.
The significant failure modes for Silver Creek Dam, based on the PFMA workshop, were summarized in
Section 5 above. Of the significant potential failure modes identified, an early dam failure warning
system would be most effective at providing warning for the overtopping failure mode described in
Section 5.4. The system could utilize criteria such as reservoir level or the rate of rise of the pool level
(or combination of the two) to provide early detection of overtopping concerns. The system could
provide notification at various thresholds to indicate how significant the conditions are. An important
elevation for overtopping concerns is Elev. 435, the top of the impervious core section of the dam. The
core section is the low permeability zone within the dam that provides the ability of the dam to
impound water. Increased seepage through the embankment fill above the impervious core zone would
be expected to occur for pool levels above El. 435 since the pool is above the top of the core. In
addition, wave action at these high pool levels may erode the crest of the dam resulting in potential
overtopping at elevations less than the current crest of the dam. Section 7 includes additional
considerations and recommendations regarding overtopping.
The other credible and significant failure modes are related to rapid erosion of the foundation material
beneath the spillway due to the loss of spillway slabs (see Sections 5.1 through 5.3). The most effective
means of detecting a developing condition for these failure modes is to regularly inspect for offsets
between adjacent slab joints and for scour in the stilling basin and on the upstream sill. The most critical
offset is the downstream slab uplifted relative the upstream slab which creates localized high pressures
that may penetrate down along the slab joint and create high uplift pressure beneath the slab. For the
upstream sill scour failure mode, described in Section 5.3, it is anticipated that progression of this failure
mode would result in increased seepage in the horizontal drains installed within the landslide deposit. A
sustained, gradual increase in seepage during spilling that is not related to rainfall would be expected to
occur. The early warning system should include automated collection of drain data from the drains
(either individually or some drains can be combined). This will require construction of weirs to collect
the flow. The City could also consider installing a new shallow piezometer to monitor development of
this failure mode. The piezometer could be installed to the left of the left spillway wall downstream of
the filter zone to a depth of approximately 10 feet below the base of the spillway with a pressure sensor
for automated monitoring. Increased uplift pressure beneath the spillway should result in increased
piezometric level in this piezometer due to hydraulic connectivity of landslide material beneath the
spillway and that screened by the piezometer. While the weirs and piezometer might provide indication
of progression of this failure mode, visual inspection of the upstream sill and clay blanket is a more
direct and effective means for detecting if erosion is occurring. Also, it is important to note that the
automated system would probably not be able to provide early detection and warning for the slab
jacking (Section 5.1) and stilling basin and downstream sill scour failure modes (Section 5.2).
In addition to the benefits described above, an automated warning system that includes data collection
from the existing piezometers and new drainage collection weirs near the toe of the dam would provide
benefit in terms of increasing the frequency of measurements from these instruments. Although the
existing piezometers were not tied to significant failure modes, they are tied to credible failure modes
11

that the team identified. Traditionally, such measurements have been taken manually once per year at
the project, although there have been gaps of time where several years have gone by without taking
measurements. Piezometric water levels and drain flow rates are important indicators of dam
performance. These instruments should be monitored consistently at least once per month (with
automation, higher frequency is easily obtained) to understand seasonal fluctuations and long term
trends. With proper maintenance, the automated system will consistently collect the data, improving
the understanding of the behavior of the dam and reducing effort to review the historical data.
Adustments to the low level outlet gates are made using manual controls located at the crest of the dam
adjacent to the right spillway training wall. Currently, operations cannot verify the low level outlet gate
position when adjustments are made. The outlet is not visible from this location in order to confirm that
the actual gate adjustments were made. The warning system should collect data for flow out of the low
level outlet so that confirmation of gate adjustments can be obtained.
An automated warning system could also be used to notify City staff when important threshold water
levels downstream of the dam are reached. Although this notification may not be tied to failure of the
dam, it may be useful for the City to know when downstream water levels reach certain elevations to
help coordinate flood response. The system could monitor downstream water levels at various
locations and could be programmed to provide notification at multiple elevations to indicate how severe
downstream flooding conditions are becoming.
7.0 CONCLUSIONS AND RECOMMENDATIONS

7.1 Conclusions
A list of major findings and understandings from the PFMA workshop is provided in Section 4. This
section provides conclusions specifically related to the proposed early warning system.

1) The PFMA workshop resulted in the identification of 4 significant and credible potential failure
modes. Three of the 4 modes were related to erosion of the spillway foundation due to either slab
jacking, erosion of the stilling basin or erosion of the upstream sill of the spillway. The fourth mode
is overtopping of the dam.
2) Failure modes related to seepage through the right abutment landslide material are considered
credible but not as significant as the 4 modes identified above. It was judged that even under high
pool conditions, such as Elev. 440 (PMF conditions), the relative increase in hydraulic gradient
compared to what the project has experienced is not enough to initiate piping in the generally
pervious, angular, basalt fragments that comprise the landslide deposit. However, it is important to
continue monitoring seepage to detect any increases that may indicate a potential developing issue
of concern.
3) With respect to the significant and credible failure modes, an automated early warning system
would be most effective at reducing risk associated with the overtopping failure mode.
12

4) For overtopping failure modes, it is anticipated that, if properly designed and maintained, an
automated early warning system may be able to provide early warning to decision makers in time to
reduce the risk associated with overtopping failure modes.
5) Initiation and progression of the failure modes associated with erosion of the spillway foundation
would be difficult to monitor with an automated monitoring system and therefore difficult to
provide advanced warning of failure by an early warning system. Monitoring horizontal drain data
and data from a new piezometer installed to the left of the left spillway wall and downstream of the
embankment filter zone could be used to monitor development of the failure mode associated with
upstream sill scour and increased seepage through the right abutment. However, visual monitoring
for spillway slab offsets and stilling basin and upstream sill erosion are more direct ways to identify
whether these failure modes are initiating.
6) For failure modes associated with spillway foundation erosion, it is anticipated that the time from
initiation to uncontrolled release of the pool could be months if the failure occurred under normal
pool conditions but days (possibly hours) under high flood (such as a 100 year event). During high
flows, unless initiation is observed directly by City staff, minimal warning will be provided to the
downstream population for spillway failure initiated failure modes regardless of whether a
warning system is installed.
7) If properly designed and maintained, the automated data collection function of the proposed early
warning system will provide a benefit of increased consistency and quality of data (including
piezometer and drain data) collected by the system. This data helps to assess the condition of the
dam and improves understanding for how the dam responds under various pool and seasonal
conditions.
8) The automated data collection function could also be used to provide confirmation of flows through
the low level outlet and frequent updates on the elevation of Silver Creek Dam downstream of the
dam to improve coordination of flood response activities associated with hydrologic events that are
not related to dam failure.

7.2 Recommendations

7.2.1 Automated Data Collection and Early Warning System Recommendations

1) The automated early warning system should include a reservoir level measurement and rain gage.
Multiple level thresholds for reservoir level and rain data should be defined as part of the design of
the early warning system. At a minimum, the thresholds should consider the elevations of the top
of the core (El. 435) and the low spot to the right of the right abutment wall (approx. El. 438). In
addition, due to the gap in the core trench on the right abutment below the training wall, the City
should anticipate the possibility of increased seepage with pools above Elev. 430 and monitor this
location. The system should provide notification of the exceedance of the thresholds to appropriate
decision‐making personnel at the City. In addition, it should collect and store data for future
reference and analysis.
13

2) The City should design and install weir collection boxes to measure the flow rate of known seeps
and drain flows, especially flows associated with the toe drains. Changes in these flow rates may
indicate changing conditions at the dam.
3) In the event that drain flows show sustained increases in flow, consider installation of a new
piezometer to the left of the left spillway wall for the purposes of monitoring piezometric pressure
within the landslide deposit that provides the foundation for the spillway. The water level in this
piezometer could be monitored with an automated data collection system with thresholds assigned
to provide notification.
4) The reliability of the existing piezometers should be evaluated as part of the design of the warning
system.
5) For the purposes of historical data collection, the automated warning system should be designed to
collect data from the piezometers and weirs to improve the consistency of data collection and
review.
6) The automated system should support storage of historical data for the purposes of reviewing
historical trends and providing reports when needed. Time history plots of data should be reviewed
at least monthly by the City and more frequently if visual observations indicate potential developing
issues of concern.
7) For the purposes of improving flood response by the City, the automated warning system should be
designed to collect data for water level at downstream locations as determined by the City.
Multiple level thresholds can be developed to provide an assessment of level of urgency.
8) The City’s Emergency Action Plan should be updated to incorporate the early warning system and to
address notification of all citizens impacted by flooding.
7.2.2 Additional Recommendations to Improve Dam Safety and Reliability

1) Perform a survey of the dam along the crest of the dam from left abutment to right abutment in
order to determine the presence of low points on the dam. If there are low spots, they should be
raised to the design elevation.
2) Perform survey of spillway and abutment walls for vertical and horizontal deflection in order to
establish a baseline for comparison with future surveys. Additional surveys should be completed at
least every 5 years. If surveys indicate movement, the results of the surveys should be reviewed by
a structural engineer for evaluation.
3) Inspect and survey the stilling basin for scour during period when spilling is not occurring (summer
low pool inspection) in order to establish a baseline for comparison with future surveys. Inspect the
stilling basin for increased scour after large flow events. Perform the inspections and surveys at
least every 5 years.
4) Inspect the spillway for vertical offsets between slabs during period when spilling is not occurring
(summer low pool inspection). The most critical offset is the downstream slab uplifted relative to
the upstream slab; this creates localized high pressures that may penetrate down along the slab
joint and create high uplift pressure. If downstream vertical offsets are present, grind the
14

downstream slab down to remove offsets. Recommend visual inspections are completed at least
once per year.
5) Visually inspect upstream impervious blanket for holes for its full length upstream (not just the
upstream sill) when possible. Visually inspect sillfor erosion at least once per year or after large spill
events.
6) Inspect the integrity of the low level outlet pipe using a Remotely Operated Vehicle (ROV).
7) Evaluate and implement options for increasing the reliability of the low level outlet hydraulic system
and for confirmation of gate opening size. To this end, the warning system should include collection
of data on the flow through the low level outlet to improve the ability to confirm gate opening size.
8) The City’s new public notification system (Everbridge) will call all City of Silverton telephone
landlines but does not include the entire contact list from the provider of the old public notification
system (Code Red). Recommend evaluating ways to update the Everbridge system contact list for all
cellular phone users to the extent possible.
9) Perform hydraulic modeling to determine the frequency and magnitude and routing of storm events
(determine relationships between flood frequency, peak flow and downstream water elevations).
The results of the modeling would help the City with decisions regarding balancing downstream
flooding impacts with dam safety risks and to improve the City’s ability to warn the population.
Once the results of the modeling are complete, it may be appropriate to modify the operations plan
for the project.
10) The City should consider making structural improvements to prevent flooding of residents at low
elevation locations (such as Silver Gardens Retirement Home).
11) The increased woody debris building up downstream of the stilling basin creates a potential for
downstream damage during future storm events. During large events, this material may become
dislodged and cause damage to bridges and property. The City should consider removing this
material every 5 years.
15

SILVER CREEK DAM
POTENTIAL FAILURE MODES ANALYSIS (PFMA) REPORT

APPENDIX A
REFERENCES
Historical Documents
 Design Memoranda/Analysis
 Design Letter to State Engineer, CH2M, 28 March 1973, (FILE 241 – D4, towards back)
 Project Information letter to Silverton Mayor, CH2MHill, 10 April 1973, (FILE 241 – D4,
towards back) – this includes info on the outlets, suggested O&M, fish facilities.
 “Detailed Project Report – Section 205 Silver Creek Dam Early Warning System,” Corps of
Engineers, August 2004.
 “Silver Creek Dam, Section 205 Preliminary Assessment, Corps of Engineers, Draft Material,
8 July 2003.
 “Silver Creek Dam Break Analysis, Final Report,” Philip Williams & Associates, Ltd., January
18, 2000. (Have entire project file).
 “Silver Creek Dam Early Warning System Preliminary Design Report,” Squier Associates, April
2002.
 Construction Control Data
 “Final Construction Report, Silver Creek Dam,” CH2MHill, 13 January 1975. (red file)
 Trip Reports and Field Inspection Reports
 Inspection trip report, CH2MHill, 26 August 1975. (FILE 241 – D4)
 Set of 16 Oregon Water Resource Department Inspection Reports between 1981 and 2010.
 “As‐Built” Drawings
 Specifications
 Project plans and specifications, March 1973.
 Construction Photographs
 Aerial photo during construction (File #319).
 Geotechnical/Geological/Geophysical Investigations
 “An Engineering Report on Subsurface Exploration and Preliminary Design Analysis for the
Construction of a Dam on Silver Creek,” CH2M, May 1966. (This investigation appears to be
at a location downstream of the actual dam site).
 “Subsurface Investigation of a Dam Site on Silver Creek,” CH2MHill, April 1973. (Includes
geologic mapping, boring logs and soil and rock descriptions at the current dam site (B) and
at a proposed dam site downstream of the current location – location A).
 Boring Logs (included in above reports)

“Living” Documents
 Operation and Maintenance (O&M) Manual
 “Suggested Inspection, Operation and Maintenance of Silver Creek Dam,” no date, signed
‘skw’.
 Emergency Action Plan (EAP)

 Water Control Manual (WCM)
‘skw’.
 ISO9000 7.11 Dam operating procedures, includes water demands and inspection checklist
A‐1

Periodic Inspection and Continuing Evaluation
 Periodic Inspections
 “Silver Creek Dam, Phase 1 Inspection Report,” National Dam Safety Program, Oregon
Water Resources Department, June 1981.
 Intermediate Inspections
o Annual Inspection – series of State Dam Safety inspections
o Incidents
 Dam seepage memo, City of Silverton, Feb 2010. Previously undetected seepage was
encountered near downstream outlet structure, a seepage monitoring point was
constructed, photos are included.
 Instrumentation Data
 “Silver Creek Dam Piezometer Installation – Contractor – CH2MHill,” 1‐19‐78, File #319. (File
contains specs and design details for ‘new’ piezometers – appears to be for piezos P3 through
P10, includes some manual plots and tables of piezo and drain data for first couple years
following construction).
 Spreadsheets and tables of piezo and drain data for 1975 through 2007.

Evaluation and Modification Reports
 Seismic Safety Evaluation
 “Seismic Stability Analysis, Silver Creek Dam,” Cornforth Consultants, 21 July 1999.
 Dam Modification Reports
 Description and construction photos of intake repair performed in 1993 – pool was lowered
‐ and description and photos of additional repairs in 2002.
 File folder with correspondence and some sketches of repair work performed as a result of
1996 flood.
 Plans and Specifications for Concrete Repairs project, July 9, 2009.
 Operation and Maintenance (O&M) Program
‘skw’.

Other Documents
 Flood Insurance Study, U.S. Department of Housing and Urban Development, Federal Insurance
Administration, September 1978 (includes flood profiles downstream for 10 year, 50 year, 100
year and 500 year floods).
 Series of untitled plans that appear to be inundation maps.
 Current rating curve (No. 9) for Silver Creek at downstream gaging station.

 Rating curve and flow data for conditions prior to dam.
 Set of plan drawings associated with getting permits for the dam.

A‐2

SILVER CREEK DAM

APPENDIX B
PFMA NOTES
PHYSICAL DATA
 Top of Dam: 440 (439.4)
 Top of Spillway Crest: 423.67 on north end curb with 424 to South.
 Outlet invert: 376.8
 Top of Core: 435.0
 Fish Ladder Pipe Invert: 421.0
 Top of Filter: 430
 Crest width = 20ft
 Base max = 340ft
 Elev. Of Base of Dam is approx. 370ft
 Max height approx. 70 ft.
 Embankment 490ft long, spillway 160ft long
 Outlet exit invert: 371.0 (from 376.8)
 Filter zone is 4 ft wide
 Core minimum 20 feet wide
 Core trench is 24ft wide to Elev 370
GEOTECH
 Core – 4‐in minus sandy, clayey silt (ML) – low plasticity
 Zone 1 – upstream and downstream silty, sandy gravel (alluvium from river bottom‐reservoir
borrow)
 Zone 2 – downstream silty rock fragments
 Structural backfill – 3‐in. minus rock
 Drainage berm – 3‐in. minus crushed rock
 Filter zone – well graded, 4ft wide, 3‐in. minus, less than 4% fines
 North (Right) Abutment – represented by B‐13
 Silt Deposit (Landslide) on top
 Top 25ft extremely permeable (fractured, loose, open, 4‐in. angular, leaked on first
filling at spillway)
 Silty talus on top of fractured basalt
 Bedrock is at Elev. 375 (80 ft down)
 South (Left) Abutment – represented by B‐12 and B‐16 (250ft downstream)
 Plastic silt, has clay to elev. 410
 Sits on fractured basalt
 B‐16 shows boulders and gravels downstream
 Bedrock/Foundation Rock
 North – landslide, porous
 Spillway – same as north abutment
 South – interflow zone at Elev. 398 (2 basalt flows)
 Liquefaction Potential – low (maybe some small isolated zones)
B‐1

 Core trench – material was compacted in the wet (had to dewater during compaction)
 OBE = 0.0875g
 MCE = 0.4g
HYDRAULICS AND HYDROLOGY
Pools
Normal pool – 424 to 424.5
Winter Oct – Apr 424 to 427
August and September < 424
May to July – 424
Aug Sep < 424
Oct to Apr > 424
50% Pool is 424
Max pool is 427
Peak Pool Ever is 430
10yr = 429, 100yr = 431, 200yr = 432, 800yr>440, PMF>440
What is tailwater?? ********Data gap********
Flows
Normal Q in river: 50 cfs @El 424 (leakage through fish ladder)
Max pool: 200 cfs @El 424 (leakage through fish ladder and spillway)
100yr 5050 cfs @El. 431, tailwater is about El. 231.
PMF – 22,682 cfs inflow, 22650 release
Flood stage/bankfull: El. 228.2, 4 miles downstream
Gage station @ Web Street takes continuous readings and sends an auto email alert when tailwater is
within 4 ft of El. 228.2
Who gets email – Andy gets it, when it’s raining Police do a check of this.
Currently no flow predictions at this project – look for rain.
B‐2

The gage is at the low spot, this area could be built up with a wall.
Next low spot is around El. 235
Sand bagging is done at Silver Gardens Family Care (El. 227)
High school is at same elevation.
First Street @ C Street at El. 234.1
Travel time to out of bank for piping failure from breach analysis is 52 min.
Overtopping – 21 minutes to leave bank@ James Street
Flow in town for 21 hours minimum (overtopping)
Treatment Plant is much further downstream but at Elev. 204. No loss of at this site (property damage).
Pool 429 is approx. equal to bankfull at tailwater of 227.
Every year if they build a 3’ tall sand bag berm there is no other flooding.
*Conclusion – 232 is flood tailwater of concern.
OPERATIONS
42‐in. gate never used, cycled once per year. Last time operated (Winter 2009) had to replace hydraulic
cylinder in the control box. It took 1 week to have this repaired. This is the only way to dewater. In the
winter this will not draw down the lake. In the Summer, it takes about 1 day to draw down the lake.
18‐in. gate is a regulating line, adjusted for Q.
INTERVENTION
******OP ISSUE*****Access – only access to embankment side is by boat (or helicopter). Therefore,
there is no ability to place emergency materials. Therefore, can’t stop a pipe or breach.
No intervention currently possible.
Currently it takes 20 to 30 minutes for a crew person to get to the site; it takes 10 minutes to get there
from the City center
B‐3

Response – Andy calls key person with cell phone or radio to activate EOC – City manager (1), Police
chief (2), Police captain (3) and Fire Department (4). Andy would provide a recommendation on action
to take.
If bad enough, a code red is given: automated call goes out to people on a Code Red list. The City pays a
company that maintains the red code list; the City has a new system (Everbridge Call System) but Code
Red company won’t provide the City with its list. The Everbridge system will automatically call all
landlines in Silverton. However, City residents won’t be contacted on cellular phones unless they signed
up for the Everbridge call system (even if they signed up originally for the old Code Red system).
Evacuation routes are in the EAP but there has been no population education on evacuation (no public
meetings).
PERFORMANCE HISTORY
First filling in 1974, seepage was observed on Nov. 7, 1974 , when the pool was at 401.5, next day water
was seen flowing out the pipe (4‐in toe drain). Downstream wet areas broadened at the right abutment
through the landslide deposit materials; city crews installed 3 horizontal surface drains. Seepage was
believed to be occurring upstream of the clay blanket but a clay layer forced flow into the filter and
embankment.
Nov. 21, 1974 to Nov. 29, 1974, 11 horizontal drains were installed to monitor seepage rates and a
downstream berm was installed. Ability to solve the problem was adversely impacted by not being able
to get access upstream of the dam.
August, 1975 ‐ Seepage was observed between Stations 3+00 and 3+60 just above the filter and
buttress layer; it was recommended to add additional filter rock. Deterioration of clay blanket was also
observed to the right of the spillway entrance channel. Weekly observation of blanket was
recommended. Fill erosion hole at upstream end of spillway. Place gravel filter material on
downstream face of dam. Filter and drainage blanket was recommended and apparently installed. It
also appears that a 12 to 15ft deep, approx. 3 foot wide cutoff trench was installed upstream of the right
end of the embankment.
No piezometer data between 1975 and 1979.
Piezometer #2 dropped off from 1975 to 1982. This piezometer is just to the left of the spillway where
the cutoff trench was installed.
Drain data is not taken frequently enough to understand what the flows are responding to.
Seepage observed at right hand side 100 ft downstream of stilling pool.
B‐4

June 1981 Inspection
Pool level was 424.4; normal pool.
No uncontrolled seepage observed; seepage only observed in the horizontal drains.
Phreatic surface has been established; primarily controlled by seepage through right abutment talus.
Conclusion – safe dam, blanket and core trench working.
Recommendation to read drains.
1993 Scotts Mills Earthquake
Earthquake was larger than OBE (operating base earthquake)
Source was 8 miles from project.
Mw = 5.6
No damage.
1993 Inlet Structure Repair
2002 Inlet Gate Repair
Mostly maintenance work, new hydraulic lines located in a concrete trench.
Between 1981 and Feb. 2010 no evidence of seepage related distress. Known seepage quantities have
been measured approximately annually; seepage has occurred where you would anticipate it. No
evidence of turbid seepage. No signs of slope movement.
Upstream and Downstream Flow Issues
1. Low spots for flooding (Silver Gardens) – *****OPS ISSUE*****close proximity, awareness
2. Flashy reservoir
3. Landslides upstream of pool in drainage basin causes upstream dam and surge (little or no
warning) – contributes to overtopping failure modes.
B‐5

BRAINSTORMING POTENTIAL FAILURE MODES
Step 1 – Identify features, locations and issues of the project for PFMs based on background
information and site history review and the site visit.
Step 2 – Of the features identified in step 1, identify which of these have PFMs of concern associated
with them.
Step 3 – Double‐check that all standard Geotech initiation mechanisms have been accounted for.
Step 4 – Rank PFMs that we are most concerned with providing warnings for. Try to address most
significant.
Definition – Documentation refers to the area left of the spillway as the right abutment. We consider
this area as embankment. The abutment (on rock/slide) is to the right of the spillway.
Identification of potential failure modes based on brainstorming.
Spillway
1. Spillway slab jacking (drain concerns) – normal flow conditions, erodible material
2. Spillway sill downstream scour on left side – normal flow conditions, erodible foundation and
concrete irregularity at the location
3. Spillway basin scour – normal flow conditions, no survey data, previous scour (is it on
rock/seam?)
4. Spillway upstream sill scour – normal flow conditions, happened before, clay blanket may not be
fully present, can lead to uplift of spillway slab
5. Spillway right upstream wall joint –
6. Spillway right downstream training wall offset/stability – normal flow, wall falls, erosion and
headcutting
7. Spillway left downstream training wall offset/stability
8. Spillway tailwater/channel capacity‐PMF conditions
9. Spillway foundation seepage/stability/settlement (potential for differential settlement due to
different foundation upstream and downstream) – combine with PFMs 1,2,5 and 6
10. Debris blockage – PMF conditions, have bumper but no log boom; consider placing boom
upstream of clay blanket.

Right Abutment
1. Surface runoff erosion – normal flow conditions, no exit for water on right abutment,
2. Abutment seepage, Integrity of upstream clay blanket, low hydraulic gradient
B‐6

3. Inadequate training of water on right abutment (upstream of spillway) – PMF flow conditions,
abutment scour.
4. Low spot adjacent to right abutment wall (ground appears to be around 438 (438.5). Scour at
high pools. Shallow embedment of wall, lose wall, this could lead to release at pool level at El.
430*****OPS ISSUE*******
5. Landslide presence and creep movement.
6. Seismic stability/liquefaction – Cornforth report
7. Seepage at right wall, no core, going through landslide deposit with rock fragments, low head,
low gradient.

Fish Ladder
1. Spillway drain on the right hand side of the ladder. Drain is a preferential seepage pathway
alongside the drain through the granular backfill around the pipe; settlement has been
observed. Normal flow conditions, slow head cut, lots of warning time.
2. Wall stability – differential movement between adjacent blocks observed. Flow in ladder is low.
3. Ball milling within the fish ladder chambers (cobbles and boulders observed). Low Q and velocity
of flow.******OP ISSUE******
4. Operability of fish ladder gates. Ladder only, no life safety consequences.*****OP ISSUE*****
Outlets/Conduit/Controls
1. Reliability of hydraulic lines and controls, including submerged hydraulic cylinder on the gate.
Happened before where the gate breaks, Normal flow, 1 in 20 chance of failure, no drawdown
capabilities, stuck open, drain pool, lose drinking water capability, *******OP ISSUE********
2. Seepage on both sides of the outlet, around seepage collars. Normal flow conditions
*********must address********
3. Conduit integrity – earthquake loading at joints, internal erosion; founded on bedrock,
reinforced, cast‐in place, seepage collars******OP ISSUE*****
4. Hydraulic capacity of the outlets, 200 cfs is max discharge (operation at PMF – could create
problems if operating during PMF. Contributing factor to other PFMs (not a PFM on its
own)*****OP ISSUE*****
5. Inability to know what the gate position is. Contributing factor to other PFMs (not a PFM on its
own)********OP ISSUE********
6. Misoperation of outlet valve – PMF conditions – does not add consequences to overtopping in
PMF *****OP ISSUE*****(not a separate PFM)
7. Seismic induced failure of the intake tower – failure of the intake does not result in overtopping
or release greater than downstream channel capacity.
B‐7

Embankment
1. Stability – static and EQ, 1981 report and 1999 Cornforth report, high freeboard (not likely to
lose due to deformation)
2. Foundation seepage and piping – seem at 398, expect seepage but not piping due to bedrock
foundation, no evidence of instability
3. Core stability/overtopping ‐ height of core about 5 feet below crest, analyzed in stability, tested
for 40 years.
4. Filter zone size, capacity, stability in EQ; designed filter, historically performed
5. Overtopping (core less than crest)‐Lack of consistent survey on vertical settlement – potential
low spots.
6. Embankment seepage and piping (11 drains that flow, 3 drains/berm/’75 trench/blanket/2
working piezometers) ‐ burrowing animals, trees and vegetation******OPS ISSUE*****
Left Abutment
1. Seepage and piping, stability of interflow zone at Elev. 398 ; stable, shouldn’t fail due to seepage
and stability, has plastic fines (same material as core).
2. Silt foundation – seepage and piping, piping at Elev. 398 seam causes internal erosion of silt,
creates a pipe in silt and uncontrolled release
Review Geotechnical of Initiation Mechanisms
Of the 28 IMs for seepage and piping, the following IMs were deemed applicable to this structure (the
others do not apply).
IM1 – cross valley differential settlement – credible, not significant based on good performance (no
evidence) and most likely to be contributing to seepage and piping failure mechanism which is covered
elsewhere, uniform abutment slope – outlet #2.
IM14 – core foundation/abutment contact at right abutment – poor compaction results in possible
seepage and piping along conduit Outlet #2 PFM (covered elsewhere).
IM17 – seepage adjacent to conduit – covered under outlet #2 PFM
IM18 –features allowing seepage into a non‐pressurized conduit – not significant because more likely to
seep along conduits
IM19 – seepage adjacent to spillway or abutment wall – all ready covered under spillway PFMs 5, 6, 7
IM20 – seepage due to crack or gap adjacent to spillway or abutment wall – all ready covered in spillway
PFMs 5,6,7
B‐8

IM22 –backward erosion and piping through a cohesionless foundation – not credible, have core trench
that cuts it off
IM27 –scour along the contact in bedrock
Note=: The PFMs were given a preliminary rating regarding credibility and significance. The following
abbreviations were used: C&S = credible and significant, C=credible, not significant and NC=not
credible. The results of the evaluations are listed below for the identified PFMs.
Spillway
1. (C&S)Spillway slab jacking (drain concerns) – normal flow conditions, erodible material
2. (C&S)Spillway sill downstream scour on left side – normal flow conditions, erodible
foundation and concrete irregularity at the location
3. (C&S)Spillway basin scour – normal flow conditions, no survey data, previous scour (is it on
rock/seam?)
4. (C&S)Spillway upstream sill scour – normal flow conditions, happened before, clay blanket
may not be fully present, can lead to uplift of spillway slab
5. (C&S)Spillway right upstream wall joint –
6. (C&S)Spillway right downstream training wall offset/stability – normal flow, wall falls,
erosion and headcutting
7. (C&S)Spillway left downstream training wall offset/stability
8. (C&S)Spillway tailwater/channel capacity‐PMF conditions
9. Spillway foundation seepage/stability/settlement (potential for differential settlement due
to different foundation upstream and downstream) – combine with PFMs 1,2,5 and 6
10. (C&S)Debris blockage – PMF conditions, have bumper but no log boom; consider placing
boom upstream of clay blanket.

Right Abutment
1. (NC)Surface runoff erosion – normal flow conditions, no exit for water on right abutment,
2. (NC)Abutment seepage, Integrity of upstream clay blanket, low hydraulic gradient
3. (NC)Inadequate training of water on right abutment (upstream of spillway) – PMF flow
conditions, abutment scour.
4. (C&S)Low spot adjacent to right abutment wall (ground appears to be around 438 (438.5).
Scour at high pools. Shallow embedment of wall, lose wall, this could lead to release at pool
level at El. 430*****OPS ISSUE*******
5. (NC)Landslide presence and creep movement.
6. (NC)Seismic stability/liquefaction – Cornforth report
7. (C)Seepage at right wall, no core, going through landslide deposit with rock fragments, low
head, low gradient.
B‐9

Fish Ladder
1. (C)Spillway drain on the right hand side of the ladder. Drain is a preferential seepage pathway
alongside the drain through the granular backfill around the pipe; settlement has been
observed. Normal flow conditions, slow head cut, lots of warning time.
2. (C)Wall stability – differential movement between adjacent blocks observed. Flow in ladder is
low.
3. (C)Ball milling within the fish ladder chambers (cobbles and boulders observed). Low Q and
velocity of flow.******OP ISSUE******
4. (C)Operability of fish ladder gates. Ladder only, no life safety consequences.*****OP
ISSUE*****
Outlets/Conduit/Controls
1. (C)Reliability of hydraulic lines and controls, including submerged hydraulic cylinder on the gate.
Happened before where the gate breaks, Normal flow, 1 in 20 chance of failure, no drawdown
capabilities, stuck open, drain pool, lose drinking water capability, *******OP ISSUE********
2. (C&S)Seepage on both sides of the outlet, around seepage collars. Normal flow conditions
*********must address********
3. (C&S)Conduit integrity – earthquake loading at joints, internal erosion; founded on bedrock,
reinforced, cast‐in place, seepage collars******OP ISSUE*****
4. Hydraulic capacity of the outlets, 200 cfs is max discharge (operation at PMF – could create
problems if operating during PMF. Contributing factor to other PFMs (not a PFM on its
own)*****OP ISSUE*****
5. Inability to know what the gate position is. Contributing factor to other PFMs (not a PFM on its
own)********OP ISSUE********
6. Misoperation of outlet valve – PMF conditions – does not add consequences to overtopping in
PMF *****OP ISSUE*****(not a separate PFM)
7. Seismic induced failure of the intake tower – failure of the intake does not result in overtopping
or release greater than downstream channel capacity.
Embankment
1. (NC)Stability – static and EQ, 1981 report and 1999 Cornforth report, high freeboard (not likely
to lose due to deformation)
2. (NC)Foundation seepage and piping – seem at 398, expect seepage but not piping due to
bedrock foundation, no evidence of instability
B‐10

3. (NC)Core stability/overtopping ‐ height of core about 5 feet below crest, analyzed in stability,
tested for 40 years.
4. (NC)Filter zone size, capacity, stability in EQ; designed filter, historically performed
5. (C&S)Overtopping (core less than crest)‐Lack of consistent survey on vertical settlement –
potential low spots.
6. (C&S)Embankment seepage and piping (11 drains that flow, 3 drains/berm/’75 trench/blanket/2
working piezometers) ‐ burrowing animals, trees and vegetation******OPS ISSUE*****
Left Abutment
1. (NC)Seepage and piping, stability of interflow zone at Elev. 398; stable, shouldn’t fail due to
seepage and stability, has plastic fines (same material as core).
2. (C&S)Silt foundation – seepage and piping, piping at Elev. 398 seam causes internal erosion of
silt, creates a pipe in silt and uncontrolled release.
Review Geotechnical of Initiation Mechanisms
Of the 28 IMs for seepage and piping, the following IMs were deemed applicable to this structure (the
others do not apply).
IM1 – (C)Cross valley differential settlement – credible, not significant based on good performance (no
evidence) and most likely to be contributing to seepage and piping failure mechanism which is covered
elsewhere, uniform abutment slope – outlet #2.
IM14 – core foundation/abutment contact at right abutment – poor compaction results in possible
seepage and piping along conduit Outlet #2 PFM (covered elsewhere).
IM17 – seepage adjacent to conduit – covered under outlet #2 PFM
IM18 –(C)features allowing seepage into a non‐pressurized conduit – not significant because more likely
to seep along conduits
IM19 – seepage adjacent to spillway or abutment wall – all ready covered under spillway PFMs 5, 6, 7
IM20 – seepage due to crack or gap adjacent to spillway or abutment wall – all ready covered in spillway
PFMs 5,6,7
IM22 –(NC)backward erosion and piping through a cohesionless foundation – not credible, have core
trench that cuts it off
IM27 – (C)Scour along the contact in bedrock
B‐11

Credible and Significant FMs
Failure modes that are credible and significant as voted by the team which should be addressed in a risk
assessment for this project
16 Significant Failure Modes
Spillway
1. (C&S)Spillway slab jacking (drain concerns) – normal flow conditions, erodible material
2. (C&S)Spillway sill downstream scour on left side – normal flow conditions, erodible foundation
and concrete irregularity at the location
3. (C&S)Spillway basin scour – normal flow conditions, no survey data, previous scour (is it on
rock/seam?)
4. (C&S)Spillway upstream sill scour – normal flow conditions, happened before, clay blanket may
not be fully present, can lead to uplift of spillway slab
5. (C&S)Spillway right upstream wall joint – C/S for greater than 100 year event
6. (C&S)Spillway right downstream training wall offset/stability – normal flow, wall falls, erosion
and headcutting
7. (C&S)Spillway left downstream training wall offset/stability – same as #7
8. (C&S)Spillway tailwater/channel capacity‐PMF conditions
9. Spillway foundation seepage/stability/settlement (potential for differential settlement due to
different foundation upstream and downstream) – combine with 1,2,5 and 6
10. (C&S)Debris blockage – PMF conditions, have bumper but no log boom; consider placing boom
upstream of clay blanket.
Right Abutment
1. (C&S)Low spot adjacent to right abutment wall (ground appears to be around 438 (438.5).
level at El. 430
Fish Ladder
(none)
Outlets
1. (C&S)Seepage on both sides of the outlet, around seepage collars. Normal flow conditions
*********must address********
2. (C&S, low likelihood)Conduit integrity – earthquake loading at joints, internal erosion; founded
on bedrock, reinforced, cast‐in place, seepage collars
B‐12

Embankment
1. (C&S)Overtopping (core less than crest)‐Lack of consistent survey on vertical settlement –
potential low spots.
2. (C&S)Embankment seepage and piping (11 drains that flow, 3 drains/berm/’75 trench/blanket/2
working piezometers) ‐ burrowing animals, trees and vegetation
Left Abutment
1. (C&S)Silt foundation – seepage and piping, piping at Elev. 398 seam causes internal erosion of
silt, creates a pipe in silt and uncontrolled release
Number of Significant Failure Modes at Normal Pool (El. 427 ‐10 yr pool and lower)
 11 at normal load
 12 at normal load and 100 year flood (so 1 at 100 year flood)
 4 at PMF
 Seismic concerns at spillway walls, conduit
Focus on what occurs under normal loading
16 Credible and Significant Failure Modes (re‐ordered)
A. Spillway left downstream training wall offset/stability ‐ normal flow, wall falls, erosion and
headcutting
B. Spillway right downstream training wall offset/stability – normal flow, wall falls, erosion and
headcutting
C. Spillway slab jacking (drain concerns) – normal flow conditions, erodible material
D. Spillway sill downstream scour on left side – normal flow conditions, erodible foundation
and concrete irregularity at the location
E. Spillway basin scour – normal flow conditions, no survey data, previous scour (is it on
rock/seam?)
F. Spillway upstream sill scour – normal flow conditions, happened before, clay blanket may
not be fully present, can lead to uplift of spillway slab
G. Spillway right upstream wall joint – C/S for greater than 100 year event
H. Spillway tailwater/channel capacity‐PMF conditions
I. Spillway foundation seepage/stability/settlement (potential for differential settlement due
to different foundation upstream and downstream) – combine with 1,2,5 and 6
J. (C&S)Debris blockage – PMF conditions, have bumper but no log boom; consider placing
boom upstream of clay blanket.
B‐13

K. (C&S)Low spot adjacent to right abutment wall (ground appears to be around 438 (438.5).
level at El. 430*****OPS ISSUE*******
L. (C&S)Seepage on both sides of the outlet, around seepage collars. Normal flow conditions
*********must address********
M. (C&S, low likelihood)Conduit integrity – earthquake loading at joints, internal erosion;
founded on bedrock, reinforced, cast‐in place, seepage collars******OP ISSUE*****
N. (C&S)Overtopping (core less than crest) ‐ Lack of consistent survey on vertical settlement –
potential low spots
O. (C&S)Embankment seepage and piping (11 drains that flow, 3 drains/berm/’75
trench/blanket/2 working piezometers) ‐ burrowing animals, trees and vegetation
P. (C&S)Silt foundation – seepage and piping, piping at Elev. 398 seam causes internal erosion
of silt, creates a pipe in silt and uncontrolled release
Develop detailed descriptions of Significant PFMs
SILVER CREEK DAM
POTENTIAL FAILURE MODE ANALYSIS
PFM # A
PFM A – LEFT SPILLWAY WALL DEFLECTION AND EMBANKMENT EROSION
( PFM # A )
Loading: Loading – Normal pool and above 427+
Description: The existing left spillway wall/training wall yields and deflects along its length, resulting in a
crack/void on the embankment side of the wall (most detrimentally at the impervious core). Pool rises above
spillway crest (the base of the wall footing) Seepage between the core and the deflected wall (upstream to
downstream) initiates erosion of the core. Core material is carried downstream past the filter zone which
fails to act as a filter. Erosion enlarges the crack. Progression through gross enlargement results in a
continuous crack to the pool and breach to the base of the wall. Erosion would continue until the pool
recedes below the crest.
Background and Evidence:
 Existing walls deflected in several locations.
 No visual evidence of gaps between the left spillway training wall and soil backfill against the wall was
observed during the site visit.
 Wall is 16 ft tall with varying thickness and reinforcement.
B‐14

 Existing wall was patched at spillway crest joint
 Joints in the wall were pre‐formed and nailed. Joints are missing/materials pulled/plucked from the
joints.
 Existing waterstops appear broken in places.
 Footing slab width varies with wall height.
 No signs of seepage currently.
 Wall has keys adjacent to the core.
 Filter downstream of the core only goes to El 430.
 Spalled concrete showed corroded reinforcement.
 Wall founded on landslide deposit.
Conditions making PFM Likely Conditions making PFM Unlikely
Or Unfavorable Factors Or Favorable Factors
 Existing walls deflected in several locations  Wall heel is 10 ft wide.
 Existing wall was patched at spillway crest  Gradients are low.
joint  Crack would be narrow @ Elev. 424
 Joints in the wall were pre‐formed and  No visual evidence of gaps between the left
nailed. Joints are missing/materials spillway training wall and soil backfill against
pulled/plucked from the joints. the wall was observed during the site visit.
 Existing waterstops appear broken in places.  Upstream shell can act as a flow limiter.
 Wall has keys adjacent to the core.
 Filter downstream of the core only goes to El

430.
 Spalled concrete showed corroded
reinforcement
 Wall founded on landslide deposit.
 Wall not designed for current earthquake.
B‐15

* ‐ Major Contributing Factors
Knowledge Gaps and uncertainty:
 No known measurement data (settlement)
 No existing calculations
 No construction photos (for wall construction and compaction of embankment fill against the wall.
 Concrete mix and quality unknown and a concern based on spalling.
Potential Risk Reduction Actions
Ability to Detect Failure Mechanisms:
 Sinkholes/visual inspection more frequently (currently 3 inspections per year by Operations/State)
 Settlement data
 Wall surveys
 Wall drain seepage changes (quantity/turbidity)
Ability to Intervene:
 Lower pool in Summer only.
 Stockpile earthwork materials, place materials in attempt to choke the failure off. (KEY)
Interim Risk Reduction Measures:
 Lower pool (takes away water supply)
 Anchor wall (coupled with excavating embankment materials to repair void/cracks).
 Inspect more frequently.

Anticipated Warning Time and Breach Width:
You can’t see this happen.
With low gradients and clay present, breach will take few days.

B‐16

SILVER CREEK DAM
PFM #J& K & N
PFM J& K & N – OVERTOPPING
( PFM # J&K & N )
Loading: Loading – Pool above 435 ft.
Description: Pool rises above the El. 435 (top of core), wave action initiates erosion of sandy/gravel in the
embankment upstream shell and crest. Material sloughing upstream and downstream decreases the crest
width/seepage path leading to overtopping. Continued overtopping leads to downcutting of the
embankment. Breach occurs, down‐cutting to the valley floor leading to uncontrolled release and
downstream consequences.
 Core only designed to El. 435.
 Historical debris blockage could raise the pool.
 Actual crest elevation varies and is unknown.
 Low spot occurs at right abutment at El. 438 would permit overtopping at lower elevations than PMF
or rest of project.
 Max pool elevation to date is 430.8 (5620 cfs) – 1996 may have been higher (no data)
 Project cannot pass the PMF
 Flashy pool
 Spillway has bumper constructed to prevent debris blockage.
 Insufficient free board
 Overtopping duration should be approx. 7 hours.
 Short fetch for wave action
 Sandy gravel on upstream face
 No forecasting data
B‐17

 Core only designed to El. 435.  Max pool elevation to date is 430.8 (5620 cfs) –
 Historical debris blockage could raise the 1996 may have been higher (no data)
pool.  Spillway has bumper constructed to prevent
 Insufficient free board debris blockage.
 Project cannot pass the PMF  Overtopping duration should be approx. 7
 Flashy pool hours.
 Actual crest elevation varies and is unknown.  Short fetch for wave action
 Low spot occurs at right abutment at El. 438  Event that would cause overtopping is
would permit overtopping at lower infrequent.
elevations than PMF or rest of project.

 No current surveying data
 Elevations unknown
 PMF unknown
 Visual observation only (impaired at night).
 Flooding
 None
 Raise low spots
 Armor the upstream face
 Armor the downstream face
B‐18



Warning time – hours
Breach time – 2 hours
We need several triggers for flood warning and overtopping
1) Trigger for 24 hour surveillance
2) Downstream levels that are critical to flooding for nuisance.
3) Downstream levels that are critical for life safety.
4) Pool elevations that create these downstream levels for overtopping.
5) Stop Prediction.
6) Communication with Public.
SILVER CREEK DAM
PFM #L
PFM# L – Seepage and piping adjacent to outlet conduit
( PFM # L )
Loading: Loading – Normal pool and above 427+
Description: Under normal pool, hydraulic gradient exists across the core zone. Seepage through the core is
designed to exit through the downstream filter zone. Due to poor compaction during construction, voids are
present adjacent to the conduit downstream of the filter zone. Sufficient gradient exists to initiate erosion
of the fines in the zone 2 embankment or conduit backfill material into the voids. Continued erosion of the
fines results in more porous rock fragments. The remaining open framework and rock permits continued
erosion of fines as it works upstream to the filter. The open backfill/zone 2 materials allows the vertical filter
drain material to erode into zone 2. Loss of filter initiates erosion of the clay core. Erosion of the core
continues by plucking material from the core and stoping. Stoping progresses toward the upstream face
through continued stoping and sloughing. Breach occurs through loss of freeboard and continued
downstream sloughing resulting in uncontrolled release of pool and consequences.
B‐19

 Seepage is presently observed at the outlet structure on both sides of the conduit.
 A pipe was installed to capture seepage on the right hand side of the outlet structure in Feb. 2010.
 The filter drain downstream of the core is designed to filter the core material and permit seepage to
exit downstream through a 4’ wide drain located at the lowest point in the foundation (old river
channel) and exits at the toe of the dam on the left side of the outlet conduit.
 Clay core is compacted against the conduit around collars placed at 10 foot centers.
 Filter was designed (well graded, 1.5‐in. to #50 sieve, less than 4% passing the No. 200 sieve)
 Clay core is sandy clayey silt.
 Zone 2 shell material is reworked landslide debris from the spillway excavation.
 Conduit backfill was hand compacted
 Zone 1 material is silty sandy gravel.
 No construction records are present, no construction photos.
 Pool has been at spillway crest for 35 years.
 Observed seepage to date is clear.
 No collection/measurement of seepage is present in this area.
 Seepage can only be observed standing at the toe of the dam which requires boat access
 Project operations staff are physically on the embankment 3 to 4 times per year.

 Seepage is presently observed at the outlet  Filter was designed (well graded, 1.5‐in. to #50
structure on both sides of the conduit. sieve, less than 4% passing the No. 200 sieve
 Seepage has increased on the right side of  The filter drain downstream of the core is
the conduit in recent years. designed to filter the core material and permit
 Clay core is compacted against the conduit seepage to exit downstream through a 4’ wide
around collars placed at 10 foot centers. drain located at the lowest point in the
 Clay core is sandy clayey silt. foundation (old river channel) and exits at the
 Conduit backfill was hand compacted toe of the dam on the left side of the outlet
 Zone 1 material is silty sandy gravel. conduit.
 Pool has been at spillway crest for 35 years.
 Observed seepage to date is clear.

B‐20

 Quality of construction of QC unknown, no photos, minimal construction records.
 Quantity and change in seepage is unknown.
 Lack of in‐situ data prevents filter compatibility evaluation.
 Other potential sources of seepage (runoff, conduit, leakage, etc) could exit at the outlet structure as well
making detection and trending difficult.
 Rate of seepage is monitored twice per year.
 Would not be seen during an inspection until the hole is big enough.
 Cloudy flow would/could only be observed with a boat ride over to the toe of the embankment.
 Project staff does target this area all ready as a known concern.

 Can only lower the pool in Summer
 No access for emergency repairs.

 Enhanced monitoring for seepage and piping.
 Collection box and weirs to separate and measure flows and measure turbidity.
 Educate community and City regarding reasoning for seepage. It should seep here. This was how the
dam was designed.

 Breach formation time – time from initiation breach to loss of dam. Once it starts to breach, the
warning time begins, it will take 4 hours from breach initiation to full breach (note: should
undertake study to determine the breach formation time).
 Warning opportunity time is 0 or negative because the dam is not monitored constantly. Pipe
can form, cloudy flow before you notice it.
SILVER CREEK DAM
PFM #D & E
B‐21

PFM# D & E – Downstream Sill and Stilling Basin Scour
( PFM # D & E )
Loading: Loading – All pools above El. 424
Description: Spillway flows initiate erosion of the basalt and spillway stilling basin concrete, continues to
erode materials, undermining the spillway concrete and foundation. Failure/undermining of the spillway
foundation at the toe results in loss of slab support. Loss of slab support permits slab failure, exposing the
talus/landslide foundation. Exposed talus is highly erodible and scour will progress. Continued loss of talus
results in failure of additional slabs and additional foundation exposure and erosion. Erosion progresses to
the upmost slab, resulting in all spillway slabs having been displaced and eroding down to basalt resulting in
uncontrolled release.

 Spillway centerline was turned 40 feet south during construction due to poor foundation conditions.
 Construction final report documents difficulty with rock anchor installation due to poor rock quality
and saturated foundation (water in anchor holes).
 The first/lowest slab is shown founded on basalt.
 Anchors are 8 foot centers on bottom slab, 10 foot centers above.
 Joints are key‐shaped with pre‐molded joint‐filler between slabs with a plastic water‐stop embedded
approximately 3‐inches below top of spillway slab (no doweling or structural tie between slabs).
 Original inspection manual is concerned about erosion of the stilling basin and recommends frequent
inspections and monitoring for erosion.
 Inspections/records are not available and inspections have not been done in recent years.
 High flows are infrequent.
 According to City, historical peak pool is Elev. 430, corresponds to 5,000 cfs based on spillway
discharge rating curve (1981 USACE Report).
 Pool was very calm immediately downstream of end of spillway then turbulent downstream of calm
area (possible deep pool).

B‐22

 Spillway was extended due to poor  The first/lowest slab is shown founded on
foundation conditions (talus) basalt.
 Construction final report documents difficulty  Spillway flows occur approximately 9 months
with rock anchor installation due to poor rock out of the year.
quality and saturated foundation (water in  High flows are infrequent.
anchor holes).
 Oct. 1975 inspection mentions scour at the
sill.
 Visual observation during the site visit
showed turbulence/flow disturbance on the
left side of the spillway at the toe.
 Foundation bedrock was described as poor to
fair quality.
 Joints are key‐shaped with pre‐molded joint‐
filler between slabs with a plastic water‐stop
embedded approximately 3‐inches below top
of spillway slab (no doweling or structural tie
between slabs).
 Original inspection manual is concerned
about erosion of the stilling basin and
recommends frequent inspections and
monitoring for erosion.
 Spillway flows occur approximately 9 months
out of the year.
 Pool was very calm immediately downstream
of end of spillway then turbulent
downstream of calm area (possible deep
pool).

 Current extent of scour is unknown.
 Unclear what is causing the flow disturbance at the left toe.
 No as‐built drawings that show spillway changes or as‐built elevations.
 Failure is underwater all year.
 Would only see this after slab moves.
 No diver or survey information
B‐23

Ability to Intervene (during failure):
 Cannot lower pool or stop spill.
 Could place rock with crane.

 Monitor for scour/survey the area.
 Armor stilling basin.

 Breach formation time – time from initiation breach to loss of dam. For normal pools – months,
for high flows greater than 100 yrs – days (maybe hours).
 Warning opportunity time is 0 or negative. You will see this after a slab has failed (maybe).
SILVER CREEK DAM
PFM #C
PFM# C – Slab Jacking
( PFM # C )
Loading: Loading – All pools above El. 424
Description: Failure modes is similar to D&E except for different initiation mechanism. Spillway flows impact
on offset in spillway slabs which result from uplift pressure or slab settlement (either under long term static
conditions or seismically induced. Landslide foundation materials are exposed initiating erosion. Erosion and
continued slab failure due to loss of support continue and progress to the pool.

B‐24

 Rock anchors were installed beneath the spillway slabs between the downstream end of the spillway
and approximately 50 feet upstream of the end. The anchors are on 8 foot centers on the bottom
slab, and 8 foot centers longitudinally and 10 foot centers transversely above the bottom slab.
 Original inspection manual is concerned about erosion of the stilling basin and recommends frequent
inspections and monitoring for erosion.
 Inspections/records are not available and inspections have not been done in recent years.
 Pool was very calm immediately downstream of end of spillway then turbulent downstream of calm
area (possible deep pool).
 No known offsets
 No data/inspection except visual flow disturbances documented during site visit.
 Drains are not inspected/cleaned.
 Drains travel underneath the slab across to a 6‐in. pipe which runs down the ladder and empties into
the ladder.
 Uplift pressure unknown.

 Construction final report documents difficulty  No known offsets
with rock anchor installation due to poor rock  Spillway flows occur approximately 9 months
quality and saturated foundation (water in out of the year.
anchor holes).  High flows are infrequent.
 Joints are key‐shaped with pre‐molded joint‐
filler between slabs with a plastic water‐stop
between slabs).
 Visual observation during the site visit
showed turbulence/flow disturbance on the
B‐25

left side of the spillway at the toe.
out of the year.
 Pool was very calm immediately downstream
of end of spillway then turbulent
downstream of calm area (possible deep
pool).
 No data/inspection except visual flow
disturbances documented during site visit.
 Drains are not inspected/cleaned.

 Presence of offsets.
 Design uplift pressures.
 Current uplift pressures.
 Effectiveness of drains.
 Failure is underwater all year.
 Would only see this after slab moves.

 Cannot lower pool or stop spill.

 At times of non‐spilling, inspect for offsets.
 Grind slabs flush.

for high flows greater than 100 yrs – days (maybe hours).
B‐26

SILVER CREEK DAM
PFM #O
PFM# O – Seepage and Piping through the Embankment ‐ Embankment foundation seepage and piping
in the landslide deposit foundation at contact with core.
( PFM # O )
Loading: Loading – All pools.
Description: An existing hydraulic gradient is sufficient near Station 4+75 to initiate silt and clay sized
particles in the landslide foundation to migrate downstream into one or more of the 11 unfiltered horizontal
drains installed after construction to deal with seepage after first filling. Erosion continues upstream or
vertically through the landslide deposit leaving a porous open‐work rock matrix. A porous openwork path is
formed which is in direct contact with the core. Erosion of the core material initiates and progresses as
impervious blanket is displaced downstream. Sinkhole forms as zone 1 materials collapse. Collapse of
upstream sinkhole/shell material allows direct connection with the pool and significant seepage/uncontrolled
release.
(After considerable discussion, the team feels this failure mode is credible where the core layer is founded on
landslide deposits and core material transport may be possible into porous landslide material. Bedrock was
treated with grout where core is underlain by bedrock.)

 Seepage on first fill near Station 4+75 was observed.
 Foundation is talus landslide deposits (silty clay rock fragments).
 11 horizontal drains and 3 surface drains installed in this area after construction.
 Core and upstream impervious blanket are thin (4’) and in direct contact with the landslide
foundation, which is untreated.
 Zone 1 material is silty, sandy, gravel (cohesionless) and serves as plug for sinkhole.
 The downstream berm was installed to address this seepage but has no effect on this failure mode.
 There is no filter zone underneath the impervious blanket.
 Drains are monitored about 3 times per year.
 Gradient for this flow path exists year‐round.
B‐27

 This load has existed for 30 + years.
 When observed, discharge from the drains has been clear.
 Hydraulic gradient is near 1 at the contact.
 Limited access to dam.

 Seepage on first fill near Station 4+75 was  When observed, discharge from the drains has
observed. been clear.
 Foundation is talus landslide deposits (silty  Zone 1 material is silty, sandy, gravel
clay rock fragments). (cohesionless) and serves as plug for
 11 horizontal drains and 3 surface drains sinkhole.
installed in this area after construction.
 Core and upstream impervious blanket are
thin (4’) and in direct contact with the
landslide foundation, which is untreated.
 There is no filter zone underneath the
impervious blanket.
 Hydraulic gradient is near 1 at the contact.
 Gradient for this flow path exists year‐round.

 True composition of landslide deposit is unknown in this area.
 Turbid flow and increased seepage in drains.
 Visual evidence of sinkholes on the embankment.
 Whirlpools near the upstream face of the dam.

 Place sand and gravel in sinkholes.
B‐28

 Monitor flow and turbidity in drains.
 Install new, filtered drains.

 Breach formation time – May not even breach, breach time would be days due to flow limiting.
Uncontrolled release is defined here as a large seep under the core. This may not result in a
true breach of the dam.
 Warning opportunity (turbidity and flow detection) time is weeks or months; is a slow process.
SILVER CREEK DAM
PFM #F
PFM# F – Scour of upstream spillway sill.
( PFM # F )
Loading: Loading – All pools above El. 424.
Description:
Keys:
1) Blanket failed.
2) Drains are overwhelmed by increased seepage due to blanket failure.
3) Increased uplift on slabs.
4) Slabs fail.
5) Erosion/scour.
6) Loss of spillway to bedrock.
Scour present on the upstream sill of the spillway permits additional flow under the spillway. The resulting
flow overtaxes the spillway drain system creating full uplift on the spillway and slabs. Increased uplift and
cycles of loading displace spillway slabs (by offset, settlement, anchorage failure, etc.) resulting in exposed
highly erodible talus foundation. Erosion progresses through continued loss of slab support and scour. Scour
progresses upstream to full pool resulting in uncontrolled release down to bedrock (El. 375).
(This failure mode is similar to C, D and E only different initiating mechanism.)
B‐29

 Anchors are 8 foot centers on bottom slab, 10 foot centers above.
 According to City, historical peak pool is Elev. 430, corresponds to 5,000 cfs based on spillway
discharge rating curve (1981 USACE Report).
 Upstream impervious blanket has historically eroded and been repaired with pit run rock.
 Each repair cycle indicates removal of impervious blanket, resulting in reduced scour but increased
flow beneath the spillway.
 Piezometer level in B‐2 , installed in the landslide deposit has been relatively constant since
installation.
 Construction final report documents difficulty  Spillway flows occur approximately 9 months
with rock anchor installation due to poor rock out of the year.
quality and saturated foundation (water in  High flows are infrequent.
anchor holes).  Piezometer level in B‐2 , installed in the
 Joints are key‐shaped with pre‐molded joint‐ landslide deposit has been relatively constant
filler between slabs with a plastic water‐stop since installation.
between slabs).
out of the year.
 Upstream impervious blanket has historically
eroded and been repaired with pit run rock.
 Each repair cycle indicates removal of
impervious blanket, resulting in reduced
scour but increased flow beneath the
spillway.

B‐30

 Condition of the blanket upstream of the sill.
 Uplift pressures.
 Slab design (designed to full uplift?).
 Visual observation of flow disturbance.
 Decreased spillway drainage downstream may indicate uplift increase.
 Drop boulders in the hole.

 Ensure spillway drains are effective.
 Ensure blanket remains impervious.
 Anchor slabs if needed.
 Add piezometer adjacent to the spillway.

for high flows greater than 100 yrs – days.
SILVER CREEK DAM
B‐31

PFM #P
PFM# P – Internal erosion of left abutment silt foundation material.
( PFM # P )
Loading: Loading – All pools.
Description:
Sufficient hydraulic gradient exists to initiate erosion of the plastic silt material into the zone 1 downstream
shell material near elevation 400 at the contact between zone 1 and silt colluvium abutment, where no filter
exists.
(After much discussion, the team decided that this failure mode is not likely due to seal of core trench into the
silt deposit at the left abutment. The only source of water producing hydraulic gradient to drive the
mechanism is surface runoff or hillside seepage which doesn’t provide sufficient gradient.)





B‐32

for high flows greater than 100 yrs – days.
SILVER CREEK DAM
PFM #M
PFM# M – Seismically Induced deformation in outlet conduit.
( PFM # M )
Loading: Loading – Seismic
Description: Seismic loading causes displacement of conduit due to material (clay, zone 2, etc) being
seismically excited at different modes. If this occurs and a gap opens in the conduit, the failure mode would
progress as PFM – L with a different probability of initiation.

 Clay core is compacted against the conduit around collars placed at 10 foot centers.
 Filter was designed (well graded, 1.5‐in. to #50 sieve, less than 4% passing the No. 200 sieve
 Clay core is sandy clayey silt.
 Zone 2 shell material is reworked landslide debris from the spillway excavation.
 Conduit backfill was hand compacted
 No construction records are present, no construction photos.
 Flow in the pipe is low (gradient is low for moving material).
 Conduit has never been TV inspected.
 Conduit was not designed for 0.4g.
 Founded on bedrock.
B‐33

 Total deformation, even @0.4g would be small.
 OBE = 0.0875g
 MCE = 0.4g
 EQ at Scott Mills was larger than OBE.

 Conduit backfill was hand compacted  Flow in the pipe is low (gradient is low for
 Conduit has never been TV inspected. moving material).
 Conduit was not designed for 0.4g.  Total deformation, even @0.4g would be
small.
 Founded on bedrock.

 Condition of conduit.
 Capacity of the conduit to handle deformation.
 Extent of deformations in a seismic event is unknown.
 Respond due to earthquake.
 Turbidity in flow from conduit.
 Close the outlet gate.
 TV inspect the outlet pipe.

 TV inspect the outlet pipe.

B‐34

 Breach formation time – time from initiation breach to loss of dam. Once it starts to breach, the
warning time begins, it will take 4 hours from breach initiation to full breach (note: should
undertake study to determine the breach formation time).
 Warning opportunity time is 0 or negative due to inability to predict earthquakes.
Reconsidering Ratings of 16 Credible and Significant Failure Modes
A. Credible, not significant. Spillway left downstream training wall offset/stability –Credible failure
mode, but not significant for normal pools/flows due to low driving gradient through the
embankment/impervious core adjacent to the crest of the spillway. This is a contributing factor
however to PFM C and PFM D. Failure of the training walls will lead to flows not trained
properly and potential increased scour of the stilling basin or the spillway slabs. Walls should be
monitored and inspected for additional movement.
B. Credible but not significant (low consequences). Spillway right downstream training wall
offset/stability – RT wall failure will expose the fish ladder. Consequences – fish and
uncontrolled flow that would continue to flow until pool lowers in summer. Even at high spills,
flow will be impacting weirs which act as baffles. No life consequences. Fish and economics.
Cannot dump sufficient water downstream to be considered an independent failure mode.
Contributing factor to PFM C and PFM D. Walls should be monitored and inspected for
additional movement.
C. Credible and Significant ‐ Spillway slab jacking (drain concerns) – normal flow conditions,
erodible material.
D. Credible and Significant ‐Spillway sill downstream scour on left side – normal flow conditions,
erodible foundation and concrete irregularity at the location.
E. Credible and Significant ‐ Combined with D ‐ Spillway basin scour – normal flow conditions, no
survey data, previous scour (is it on rock/seam?)
F. Credible and Significant ‐ Similar to C,D,E in consequences – (Spillway upstream sill scour –
normal flow conditions, happened before, clay blanket may not be fully present, can lead to
uplift of spillway slab
G. Credible, not significant for Normal Pools ‐ Spillway right upstream wall joint – C/S for greater
than 100 year event ‐ High pool, load up training wall, loading up the landslide deposit.
Insufficient hydraulic gradient to move material downstream, wall should not fall, just leak (this
is right wall issue).
H. Credible, not significant ‐ Spillway tailwater/channel capacity‐PMF conditions ‐ If flow is high,
debris will move, no backwater effects. No routing or channel capacity evaluation done. Existing
channel may not have been designed for new 100 year flood. Recommend analysis be
performed for whether 100yr, 500yr etc. floods stay within existing banks etc. This is critical to
property damage, evacuation routes, population awareness etc.
B‐35

I. Credible, not significant ‐ Spillway foundation seepage/stability/settlement (potential for
differential settlement due to different foundation upstream and downstream) – combine with
1,2,5 and 6 ‐ Spillway settlement and stability under seismic conditions. The 1999 stability
analysis did not look at the spillway talus foundation. However, the report did look at slope
stability with embankment material having frictional strength of 36 degrees. The team believes
liquefaction and slope stability/sliding failure of the talus foundation is unlikely due to the
composition of the talus (high friction and the shallow slope). Settlement, while possible, was or
is addressed in PFM C which is slab jacking.
J. Credible and Significant ‐ combine with K and N Debris blockage – PMF conditions, have bumper
but no log boom; consider placing boom upstream of clay blanket. Will lead to overtopping at
lesser flooding events – more, frequent flood overtopping.
K. Credible and Significant ‐ SEE J and N ‐ Low spot adjacent to right abutment wall (ground
appears to be around 438 (438.5). Scour at high pools. Shallow embedment of wall, lose wall,
this could lead to release at pool level at El. 430*****OPS ISSUE*******
L. Credible, not significant ‐ Mechanism considered unlikely with respect to necessary voids
forming. Seepage on both sides of the outlet, around seepage collars. Normal flow conditions
Not significant – limited flow/erosion capacity based on dam construction materials.
M. Credible, not significant. Conduit integrity – earthquake loading at joints, internal erosion;
founded on bedrock, reinforced, cast‐in place, seepage collars Not significant flow, inspect
conduit and reevaluate. If conduit has structural integrity issues could collapse and result in
overtopping/crest deformation.
N. Credible and Significant ‐ SEE J AND K(C&S)Overtopping (core less than crest) ‐ Lack of consistent
survey on vertical settlement – potential low spots. Significant and can be tied to early warning
system.
O. Credible, not significant ‐ Mode is for where core is on top of landslide (4 foot) blanket at
minimum thickness – may not breach, just uncontrolled seepage ‐(C&S)Embankment seepage
and piping (11 drains that flow, 3 drains/berm/’75 trench/blanket/2 working piezometers) ‐
burrowing animals, trees and vegetation.
P. Considered credible not significant ‐ Mechanism unlikely (no source of water to drive erosion)
Left abutment, Silt foundation – seepage and piping, piping at Elev. 398 seam causes internal
erosion of silt, creates a pipe in silt and uncontrolled release Only could occur at high pools and
nowhere for material to pipe to.
Major Findings and Understandings
1. The Dam appears to be well maintained and project operations staff is maintaining the project
to the best of their abilities.
2. A variety of documents were available for review for the PFMA. In general, there was a lack of
construction photos associated with original construction (one aerial photo was available from
original construction).
B‐36

3. General access to the right side of the project is by car. Access to the right abutment of the
project site (including the right spillway training wall and fish ladder) can be gained on foot.
Access to the embankment, low level outlet and left abutment of the dam requires a boat.
4. Under normal operations, pool rises to spillway crest (El. 424 feet) for about nine months of
every year. Pool of record is about El. 430 feet (6 feet over the crest) .
5. Downstream flooding occurs every year at Silver Gardens Retirement Home, requiring sand
bagging every year. This will continue unless structural measures are taken to prevent flooding.
6. There is no stream gauging upstream of the dam or forecast data available at the project. As a
7. There are no weirs to measure seepage present at this project. The city currently manually
8. There is no seepage collection/measuring present on the left abutment. Because of the
historical seepage concerns, seepage was extensively discussed. For this PFMA, general seepage
was broken down into seepage leading to specific failure mechanisms (i.e. increase seepage
causes an increase uplift pressure beneath the spillway slabs leading to failure of spillway slabs
allowing erosion of the underlying foundation materials to breach the dam).
9. Outlet capacity and reliability are keys to safe project operations. The project cannot dewater
quickly if needed for an emergency (except in summer). In addition, there are reliability
concerns based on operation history (This needs to be addressed by the City – will the outlet
work when needed?).
10. Inspectors/City Staff cannot view or inspect the left side of the Dam (embankment, spillway and
11. City staff cannot intervene to stop dam failure without the use of a helicopter or boat/barge.
difficult.
12. The current Emergency Action Plan (EAP) does not allow for full contact with the population
(how is the population of the city of Silverton notified?). The call list for the current emergency
public notifications system (Everbridge Call System) is not complete. According to City
Operations Staff, the suppliers of the old notification system (Code Red) will not release the full
list to the City. The Everbridge system will automatically call all landlines in Silverton. However,
City residents won’t be contacted on cellular phones unless they have signed up for the
Everbridge call system (even if they signed up originally for the old Red Code system).
13. The Project Operations plan needs trigger thresholds for inspection (when to go to 24‐hour
surveillance, etc.) and warning. Critical elevations need to be established for this project in
order to ensure adequate inspection and early warning.
14. Known foundation features will continue to seep (basalt interflow zone @ 398, landslide talus
on right abutment). However, these areas are not prone to piping and dam failure.
15. Visible signs of distress are present at the dam (the spillway training walls are tilted, stilling
basin scour is apparent at the site, new and/or increased seepage has been identified in recent
years).
B‐37

16. The City reports that seepage has increased in recent years (left and right side of outlet works,
new pipe installed in Feb 2010), but there is no way for the City to quantify or measure to verify
if seepage is increasing due to a lack of a weir or other measurement system that could be used
to quantify this seepage.
17. The City inspects the dam twice per year and the State inspects the dam approximately once per
18. Stilling basin scour is evident and has been previously reported in post construction inspections.
There is no record of scour inspection or monitoring by the City. . Original recommended
operations and maintenance document indicates that the stilling basin is prone to scour and
should be frequently inspected. The inspections will require diverting water from the low level
outlet and fish ladder and pumping from the stilling basin during periods of no flow through the
spillway; or use of an ROV.
19. Power at the project is City supplied and is only currently available on the right side of the Dam.
No backup power, such as an emergency generator, is available at the site. Some method of
20. Some of the piezometers are currently non‐functional and some of the piezometer standpipes
are small in diameter. Small diameter piezometers may prove difficult to automate.
21. Operations personnel have no means of verifying the low level outlet gate position (42‐in. or 18‐
in outlet) – operations cannot see the outlet from the control stand located on the left bank.
22. Outlet operation during high flows (open or closed) is not defined in the operations manual.
The outlet may not be structurally sound at high pools when the gate is open due to the
potential for cavitation. In addition, the air intake for the outlet may be inundated in high flow
events which will result in surge flow.
23. Low spots are present at the right abutment. This means overtopping may occur at lower
elevations.
24. No current hydraulic modeling/routing has been performed in order to define thresholds (100 yr
flood, PFM, etc.), pools and tailwaters. If modeling data was available, the ability to set
meaningful thresholds for an early warning system would be improved.
25. While the top of the embankment is designed at elevation 440, the clay core within the
above the clay core should be anticipated with pool levels greater than El. 435. In addition, the
lack of a clay core extending to the top of the dam may result in overtopping occurring at
elevations below elevation 440 (i.e. overtopping may begin before the pool reaches the top of
the dam due to wave/wind erosion of the embankment). The potential for overtopping to begin
at a lower elevation should be considered when defining trigger points for an early warning
system.
26. The City was not able to produce current survey data for true height of embankment crest, the
tops of walls or abutments. Settlement of the embankment or other low spots may result in
overtopping at lower elevations than anticipated. Knowing the location of low spots is
important for determining locations of stocking emergency materials and to focus flood fighting
efforts.
B‐38

27. Based on the results of analysis performed to date by Cornforth Consultants, the embankment
and abutments are expected to be adequately stable under seismic conditions and liquefaction
28. No seismic analysis of the concrete spillway or spillway wall has been performed. It is
anticipated that loss of the wall may lead to failure of the embankment. (Walls designed for
0.05g, current MCE from USGS = 0. 4g).
29. There appears to be a gap in the core trench on the right abutment below the training wall. The
presence of a gap at this location may result in seepage at pools above Elev. 430. The City
should anticipate the possibility of increased seepage with pools above Elev. 430 and monitor
this location.
30. Intervention materials (sand bags, rip rap) are not currently stockpiled on the site.
31. There are residence locations within very short travel times downstream of the dam. Warning
time would not be possible for some failure modes.
32. Loss of pool in summer months will eliminate the downstream water supply and force the City
to go to water conservation.
33. A log boom was installed at the end of construction at the upper end of the pool but is no longer
34. Large vegetation debris is expected to continue to collect downstream of the spillway below the
35. Project staff was unaware of the presence of 4 foot filter drain that runs from the filter zone to
the left of the outlet pipe and ends downstream immediately to the left of the outlet structure
(this may or may not exist since as‐built drawings do not show it but it is discussed in CH2MHill
correspondence). If it exists, this drain may contribute to seepage exiting near the low level
outlet structure at the toe of the dam.
36. There are no records of conduit inspection/ROV inside the outlet pipe. The pipe should be
inspected for structural integrity.
37. Zone 1 materials on the upstream shell of the embankment consist of well graded silty, sandy
38. The upstream impervious blanket should be inspected for holes for its full length upstream (not
just at the sill).
39. The low level conduit passes through the dam at the base of the embankment. It is founded on
B‐39

SILVER CREEK DAM

APPENDIX C
SITE VISIT NOTES
Background
A site visit was conducted on Tuesday, 11 January, 2011 in order to observe and familiarize the PFMA
team with the site conditions. The purpose of this memorandum is to summarize the observations
made during the site visit. The following people were on site to participate in the site visit.
 Andy Peters (City of Silverton)
 Travis Adams (USACE, Structures, PFMA, Facilitator)
 Matthew Craig (USACE, Dam Safety, PFMA Co‐facilitator )
 David Scofield (USACE, Geotechnical)
 Thareth Yin (USACE, Hydraulic Design)
 Alison Burcham (USACE, Project Manager)
Access
The right abutment of the dam can be accessed by car. The right side of the spillway, fish ladder and
right abutment can be accessed and observed from the right abutment area. A boat ramp permits
access to the left side of the spillway, embankment and left abutment by boat. Crews access the
embankment and left side of the project with a boat. A small boat was used to gain access to the left
side of the project during the site visit.
Right Abutment Area
A wall extends from the right side spillway and extends approximately 35 feet to the right (Photo 1).
The top of the wall is at Elevation 440 according to the project drawings. A low spot approximately 1.5
to 2 feet below the wall was observed immediately to the right of the wall. It was observed that this low
spot could allow overtopping at pools less than what would occur if the ground elevation is at 440, same
as the wall. City crews repaired the concrete wall at one of its joints.
Erosion was observed in the backfill behind the upstream end of the short training wall upstream of the
right spillway wall (Photo 2). The purpose of this wall is to direct flow through the spillway away from
the spillway walls. It is anticipated that, without repairing the erosion behind the wall, increased erosion
may occur behind the wall during high flow events and potential and impair the function of the wall to
direct flow into the spillway.
No irregular surface features such as sloughing, bulging or other features were observed.
Fish Ladder Structure
C‐1

Evidence of relative lateral movement between wall segments was observed on both sides of the
spillway structure. Damage was most apparent on the right side (Photo 3). It was also observed that
the ground level immediately to the right of the right fish ladder wall was lower than the surrounding
ground. This may be an indication of settlement of granular backfill around a 6 in. drain that runs to the
right of the right fish ladder wall along its length. This drain discharges into the fish ladder structure
near the downstream end of the spillway structure. At the time of site visit, no water was flowing from
the drain into the fish ladder structure.
Cobbles and boulders were observed within the chambers of the fish ladder.
Stilling Basin
The stilling basin downstream of the spillway could not be observed; tailwater level was too high to
allow observation of the extent of scour in the spillway. It was noted that the tailwater appeared to be
relatively calm immediately downstream of the spillway (Photo 4). Such conditions may suggest that
some scouring downstream has developed, resulting in a deeper pool immediately downstream and less
turbulence near the water surface.
Some tree and vegetation debris has built up over time downstream o f the stilling basin (Photo 5). The
vegetation is likely reducing the overall channel capacity downstream of the dam.
Spillway
At the time of the site visit, water was spilling over the spillway; it appeared that the pool was slightly
above the spillway crest. Some water disruption was observed in a couple of locations; one near the
upper end of the spillway and the other near the downstream end of the spillway adjacent to the left
spillway wall.
On the left spillway wall, lateral offsets were observed between adjacent wall segments (Photo 6).
Sufficient relative lateral deformation may result in failed water stops and a pathway for water from the
spillway to flow into the backfill behind the wall.
Near the crest of the dam, concrete wall repair had previously been performed by City crews near one
of the wall joints (Photo 6). Spalling had occurred at this location previously.
Embankment
No obvious signs of embankment instability were observed (no slumps or large scale ground surface
irregularities). During the time of the inspection, seepage was observed from some of the 11 horizontal
drains that had been installed to capture seepage beneath the right end of the earth embankment after
C‐2

construction (Photo 7). On first filling, seepage had been observed on the downstream face of the dam,
especially on the right side of the embankment. The extent of the seepage was such that a downstream
buttress and a series of horizontal drains were installed to collect seepage from beneath the
embankment. Seepage was also observed on either side of the outlet structure near the toe of the dam.
At this location in February 2010, City crews installed a short collection pipe into the embankment on
the right side of the outlet structure in order to provide a means for obtaining a representative
measurement of seepage at this location (Photo 8). It is expected that the collection pipe isn’t collecting
all seepage at this location but that changes in seepage conditions would be reflected in the rate of flow
out of the pipe. There is some uncertainty regarding how long seepage has been flowing at this
location. It is possible that flow has increased here recently.
One of the piezometers was observed near the crest of the embankment. The small size of the
piezometer standpipe (drawings indicate the standpipe is ¾‐in PVC pipe) may limit options for selecting
sensors that can be suspended in the standpipes as part of the instrumentation for the proposed early
warning system.

C‐3

Photo 1. Right Abutment Wall
Photo 2. Right Abutment Training Wall
C‐4

Photo 3. Deformation at Right Fish Ladder Wall
C‐5

Photo 4. Relatively Calm Tailrace Water Conditions
Photo 5. Vegetation Debris in the Tailrace Area
C‐6

Photo 6. Relative Lateral Deformation, Left Spillway Wall
Photo 7. Seepage from Horizontal Drains
C‐7

Photo 8. Seepage Around Outlet Structure

C‐8

SILVER CREEK DAM

APPENDIX D
DRAWINGS

Silver Creek Dam Potential Failure Modes Analysis (PFMA)

Загружено:

Сведения о документе

Исходное описание:

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Silver Creek Dam Potential Failure Modes Analysis (PFMA)

Загружено:

Авторское право:

Доступные форматы

Final Report

Silver Creek Dam Potential Failure Modes

Figure 1. Site Plan

Вам также может понравиться